brain research 1580 (2014) 1–7

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Editorial Evolutionary perspectives on the role of in human social behavior, social cognition and psychopathology

The articles in this Special Edition summarize the latest 2013; Winslow et al., 2003) and the rapidly accumulating research on oxytocin (OT), and often arginine number of reports of OT influences (mostly positive) on (AVP), in human parental behavior, other social behaviors, human trust, cooperation and social cognition (e.g., see social cognition, social information processing in the brain, as Meyer-Lindenberg et al., 2011). well as developmental disorders and psychopathology. These During early placental mammalian evolution, OT was also articles aspire to not only review progress in these areas but coopted to produce affective changes vital to the success of to identify the limitations of what have been done and also to maternal behavior, e.g., reduced fear/ (see Febo and articulate high priority directions for future research. The Ferris; MacDonald and Feifel, both this volume; Neumann relevance of the emerging evidence for understanding the et al., 2000; Figueira et al., 2008), which enabled suppression psychobiology of mental illness is emphasized. To set the of newborn-directed during parturition (McCarthy, stage, here we discuss how OT in the central 1990; McCarthy et al., 1986) and regulation of intruder- (CNS) might have mediated the critical advances in social directed aggression during (Bosch et al., 2005; behavior during the evolution of placental mammals. Consiglio and Lucion, 1996; Consiglio et al., 2005; Giovenardi Selection for avid and sustained maternal behavior was et al., 1998; Lubin et al., 2003) AVP (also selected from earlier critical for the successful evolution of placental mammals. nonapeptides during the evolution of mammals) appears to OT was selected from phylogenetically earlier nonapeptides play an important role in activating and sustaining maternal to enable the unique features of placental mammalian behavior (Bosch and Neumann, 2008; Pedersen et al., 1994) reproduction: birth of neonates after in utero fetal develop- and regulating affective changes and increases in intruder ment (parturition); delivery into the mouths of suckling aggression associated with the onset of maternal behavior infants of high quality nutrition produced within the (see Febo and Ferris, this volume). These points are relevant mother's body by the larger process of lactation (milk ejec- to another theme that is central to this Special Issue: from tion); and activation of maternal nurturing of newborns. In all early on in placental mammalian evolution, OT and AVP were mammalian species tested to date, OT appears to facilitate selected for important roles in emotion regulation necessary the postpartum initiation of maternal behavior (Pedersen, for successful social behavior. 2013). In sub-primate mammals, this involves, in concert The evolution of avid and sustained maternal behavior as with and dependent upon the reproductive condi- well as other unique aspects of placental mammalian repro- tions that promote the onset of parturition, OT-initiated duction effectively eliminated long-standing barriers to motivation to exhibit hard-wired, species-specific sets of development of larger and more complex brains and, even- offspring-directed care-taking behaviors. To restate this point tually, higher intelligence. In utero development and mater- in a more general way that is a central theme of this Special nal protection during maturation as well as a uniquely rich Issue: with the evolution of placental mammals, OT was and reliable source of nutrition, i.e., milk, substantially selected to activate brain systems promoting maternal beha- increased the percentage of offspring that survived to repro- vior, a sustained and prosocial motivational state. Other duce. Therefore, numbers of offspring required for reproduc- social attachments and social behaviors that were selected tive fitness decreased allowing each offspring to be larger and later during mammalian evolution appear to be at least endowed with a bigger brain. In addition, maternal protection partially based on OT (and AVP) neural systems that first during the lactation period allowed further brain growth and emerged during the evolution of maternal behavior. Support- development to occur before offspring had to fend for ing this contention are the OT and AVP dependence of pair- themselves. This also created a malleable period during bond formation in the monogamous prairie vole (Young and which epigenetic influences from the mother, siblings and Wang, 2004) and marmosets (Smith et al., 2010), studies the environment influenced brain development in survival- indicating OT regulation of social interactions in rhesus enhancing ways (McGowan and Szyf, 2010; Champagne, macaques (Chang et al., 2012; Parr et al., 2013; Ebitz et al., 2012). Selection for mental abilities that increased the 2 Editorial / brain research 1580 (2014) 1–7

effectiveness of maternal behavior, and later paternal beha- contexts and intrinsic social orientations. For example, ICV vior, in enhancing the survival of offspring to reproductive administration of OT in male squirrel monkeys increases age may have also spurred on the evolution of more complex associative behaviors in low status males but increases brains and greater intelligence. sexual assertiveness in higher status males, upon exposures In many rodent species (e.g., rats, mice), the strong, OT- to female monkeys (Winslow and Insel, 1991). Furthermore, dependent maternal motivation that is activated at parturi- increasing OT levels in the CNS via OT inhalation in rhesus tion is directed towards newborns in general. Mothers in macaques promotes either other-regarding or self-regarding these species do not bond to individual offspring (Pedersen, behaviors depending on social decision contexts (i.e., what 2013). OT appears to have been centrally involved in the next the available options are concerning self and others) (Chang evolutionary leap in prosocial behavior in mammals; forma- et al., 2012), while controlling for the state of social vigilance tion of selective attachments to specific individuals. The toward specific social stimuli (Parr et al., 2013; Ebitz et al., formation of selective bonds requires the ability to learn the 2013). Taken together, these findings endorse the idea that OT identity of important conspecifics (e.g., offspring, mothers, interacts closely with the neural systems involved in social mating partners). In sub-primate mammals, individual iden- perception and decision-making. Chang and Platt (this issue) tification is based largely on learning the unique olfactory review selected studies of OT and social behavior in non- cues of others. Oxytocin was coopted during the evolution of human primates, focusing on the interplay between social selective social bond formation to facilitate acquisition of the motivation and social vigilance for promoting social beha- memory of odor cues of other individuals. For example, OT viors. The emergence of OT research in non-human primates release in the olfactory bulbs is necessary for ewes to encode (see Chang and Platt; Evans et al., this volume) provides a the memory of the specific odors of their newborn lambs in platform for investigating the neurobiology of central OT for the immediate (Kendrick et al., 1997). The shaping complex social cognition. Importantly, a non-human capacity to form selective, olfactory-based bonds may have primate model can further help test the efficacy and safety of emerged from the ability earlier in evolution to form tran- long-term OT-based therapies in the same subjects by sys- sient memories of the odors of novel conspecifics. For tematically monitoring neurophysiological and other physio- example, rats and mice are able to remember the unique logical and behavioral changes (see Chang and Platt, this odor of a novel individual for approximately 1–2 h. Experi- volume). ments with OT gene knockout mice or central administration In this volume, Febo and Ferris summarize their pioneer- of OT antagonists have demonstrated that formation of ing functional magnetic resonance imaging (fMRI) studies in transient social memory is entirely OT dependent (Ferguson awake rats, including nursing mothers, and Swain et al. et al., 2000; Takayanagi et al., 2005). These findings indicate provide a review including the latest fMRI studies of brain that OT has played a significant role in social cognition since activation in human parents elicited by visual and auditory early in mammalian evolution. Evidence for extensive invol- stimuli from infants. Both show the remarkably complex vement of OT in human social cognition is another major neural processing associated with parental behavior toward theme developed in many of the articles in this Special Issue. the lower end and the peak of mammalian evolution. Many During the evolution of primates, maternal and other brain areas activated by nursing stimuli in rat mothers and social behaviors have become less dependent on specific infant visual and auditory stimuli in human parents are sex and olfactory cues. The neurobiological control analogous, specifically areas involved in sensory processing, of social interactions has evolved from hard-wired behavioral mobilization of hard-wired, instinctive nurturing behaviors programs reflexively triggered by specific sensory cues to (referred to as reflexive/instrumental caring responses by more flexible and complex processing and integrating of cues Swain et al.), and emotion regulation. Febo and Ferris surmise from multiple sensory modalities. These evolutionary from evidence that many of these brain areas contain high changes in the psychobiology of social relationships in concentrations of OT receptors in rats and are activated by primates have certainly been facilitated by, and perhaps ICV administration of OT as well as that central OT release contributed to the selection for, expansion of non-olfactory during parturition may produce the coordinated activation of regions of the cortex and advances in intelligence. OT still these many brain areas during the onset of rat maternal clearly plays a crucial role in primate maternal behaviors. behavior, a conclusion that is supported by extensive studies Intracerebroventricular (ICV) infusion of OT in nulliparous in sheep (Kendrick, 2000). In humans, Swain et al. point out rhesus macaques increases their interests in infants mea- that these areas are most prominently activated by stimuli sured by looking, touching, maintaining proximity, and lip- from very young infants whose behavior is limited to com- smacking (Holman and Goy, 1995). Furthermore, peripheral municating basic needs (e.g., hunger, cold, discomfort). How- administration of OT antagonist delivered to the limbic ever, the more interactive stimuli from older human infants regions of the CNS substantially reduces nulliparous female activate higher cortical regions and cortico-limbic connec- macaque's interest in infant and sexual behavior (Boccia tions implicated in mentalization, and Theory of et al., 2007). Using non-human primates also permits unique Mind which may enable human parents to provide more investigations into the role of OT in mediating more complex attuned and sensitive responses to their infants' social cues social behaviors and social cognition than maternal beha- that are critical for the development of secure attachment in viors and pair-bonding. Converging evidence from the studies their offspring. However, little is known about relationships examining the role OT in complex social cognition in non- between central OT systems and thebrainactivationpatternin human primates as well as in humans (Bartz et al., 2011) human parents, in part because of the lack of reliable methods suggests that the effects of OT are critically gated by social to locate OT receptors in the human and non-human primate Editorial / brain research 1580 (2014) 1–7 3

brain. However, Pedersen (this issue) summarizes a recent infant distress and the incentive salience of infant laughter. immunohistochemical study (Boccia et al., 2013)thathas In light of the functional disconnect between OT effects on identified OT receptors in limbic and some cortical brain sites brain activity and infant face emotion recognition in the (unfortunately the study did not examine prefrontal and current Voorhuis et al. article, behavioral/emotional tests temporal cortices) that are activated in human parents by might be critical for interpretation of fMRI studies that look infant visual and auditory stimuli (the cover of this Special at OT-mediated modulations in hemodynamic activity. Issue illustrates OT immunostaining in the human Studies in which OT is administered by the intranasal visualized in the Boccia et al., 2013 study). route have enabled rapid advance in our understanding of OT To date, most studies of OT and parental behavior in regulation of human social behavior and social cognition as humans have examined concentration relationships well as the psychotherapeutic potential of OT. Initial, indirect with parent–infant interactions. Apter-Levi et al. (this evidence that intranasal OT penetrates the CNS was provided volume) summarize the considerable body of evidence, much by Born et al. (2002) when they demonstrated that intranasal of it from their own group, that OT levels in human parents AVP (only two amino acids different from OT) crosses the and their children are inter-related and correspond to the blood–brain-barrier. Evidence that OT administered by the degree of synchrony of their interactions. They then report intranasal route enters the CNS has more recently been the first evidence that parent AVP levels are related to reported in rats (Neumann et al., 2013) and monkeys (Chang stimulatory contact with their infant, attention to et al., 2012). Early on, intranasal OT was reported to have a objects, and efforts to increase object salience following number of prosocial effects in human subjects including infant social gaze. These novel findings add further to the increasing interpersonal trust (Kosfeld et al., 2005; bio-behavior synchrony model of parent–infant interactions Baumgartner et al., 2008; Mikolajczak et al., 2010); cooperation that has been developed by this group (Atzil et al., 2012). Kim (Andari et al., 2010) and eye contact (Andari et al., 2010; et al. (in this Special Issue) report the first study to examine Guastella et al., 2008) as well as improving performance in the relationships between OT levels and maternal responses domains of social cognition such as face emotion recognition, during conditions that are stressful to their infants (a mod- Theory of Mind, perception of the trustworthiness of faces, ified still-face procedure). They found that the duration of and empathy (Domes et al., 2007; Fischer-Shofty et al., 2010; maternal gaze toward the distressed infant, a key index of Guastella et al., 2010; Hurlemann et al., 2010; Petrovic et al., sensitive parenting, is directly related to maternal release of 2008; Theodoridou et al., 2009). Kanat et al. (this volume) OT. Elmadih et al. (this volume) employ the novel strategy of review and critique the burgeoning number of fMRI studies of comparing OT levels between mothers who exhibit high intranasal OT effects on brain region and circuit responses to sensitivity in their interactions with their infants and social stimuli. Most consistently, OT administration has mothers with low sensitivity. Surprisingly, OT levels were affected social stimuli influences on the activity of the significantly higher in low sensitivity mothers, a finding that amygdala and associated regions of the temporal and pre- contrasts with positive relationships between OT levels and frontal cortices as well as their connectivity. However, the sensitive parenting found in previous studies. Citing reports specific regions affected and the directionality of effects are that higher OT levels are associated with in social strongly influenced by differences in social stimuli, gender, relationships or disturbed parental relationships in child- reproductive state in women, and attachment style as well as hood, Elmadih et al. hypothesize that low sensitivity mother variations in the oxytocin receptor gene. may have greater stress responses to the demands of caring This Special Issue includes several articles that provide for their infants and/or experienced deficient parenting early important additions to our understanding of the complexity in life. Elmadih et al. conclude that mothers' stress reactivity of OT effects on human social behavior and social cognition. and attachment history are core variables that should be Our understanding of the often contrasting effects of OT in assessed in studies of OT relationships with parent–infant different conditions and the wide range of variables that interactions. appear to influence those effects will require a great deal of Van IJzendoorn, Bakermans-Kranenburg and colleagues future research. Evans et al. (this volume) take on the task of have pioneered fMRI studies of the effects of intranasal OT comparing the many reports of intranasal OT effects on on responses to infant stimuli. In this issue, they (Voorhuis human social behavior and social cognition. Initially they et al.) report a puzzling finding that intranasal OT adminis- examine the controversy about whether OT administered by tration in women increased activity in temporal and frontal the intranasal route penetrates the CNS. The authors then cortical sites but at the same time decreased accurate inter- view and critique the inconsistencies in the rapidly increas- pretation of images of infant facial expressions. Their results ing literature on intranasal OT effects. The notion that OT is contrast with findings that intranasal OT increases accuracy always “prosocial” is refuted. Striking gender differences in of emotion recognition when viewing images of adult faces the effects of OT are emphasized. The authors discuss (Domes et al., 2007; Guastella et al., 2010). The Van IJzendoorn suggestions by some investigators that a general effect of and Bakermans-Kranenburg group previously found that OT OT such as anxiety reduction or enhancement of the salience administration to nulliparous women decreased amygdala of social stimuli might account for the wide range of social activation and increased activity in emotion regulation sites facilitating influences of OT that have been reported. in responses to infant crying or laughter (Riem et al., 2011, De Dreu and colleagues have published a series of ground- 2012). These effects were interpreted as evidence that OT breaking studies demonstrating that OT has strikingly con- improves responses to infant's emotional states by decreas- trasting effects on attitudes and behavior directed toward ing negative emotional arousal and increasing empathy for simulated in-group and out-group members. They found that 4 Editorial / brain research 1580 (2014) 1–7

intranasal OT increases positive attitudes and empathy only that the distinct neurobiological abnormalities of these four for in-group members, enhances agreement more with in- disparate neurodevelopmental disorders may all lead to group members and promotes greater in-group cooperation dysregulation of OT control of social behavior. when competing with an out-group. Ten Velden (this issue) Virtually all psychiatric disorders are associated with summarizes their earlier work and reports the results of this some degree of impairment of social function although group's latest study. In a simulated poker game situation, profound deficits are primarily associated with severe and they found that subjects who received intranasal OT with- persistent mental illnesses such as psychotic disorders. drew from betting more, even when they had a strong hand, Numerous studies conducted in animal models of psychiatric if they were competing with an in-group member but not disorders have found that OT may have therapeutic potential when competing with an out-group member. The body of in a wide spectrum of disorders including anxiety, depres- work from De Dreu and colleagues powerfully demonstrates sion, psychosis and addiction (Macdonald and Feifel, 2013; that OT effects on human social behavior are situation Meyer-Lindenberg et al., 2011). These considerations as well dependent and could exacerbate conflict between groups. as ample evidence that intranasal OT treatment exerts effects Blandon-Gitlan et al. (this volume) report the fascinating in the CNS have led a few investigators to conduct rando- finding that intranasal OT treatment eliminates same race mized clinical trials (RCTs) in some psychiatric disorders (as bias in face recognition memory, a phenomenon that has well as neurodevelopmental disorders—see Anagnostou been strongly validated in previous studies. In contrast to et al., in this issue). Pedersen (this issue) emphasizes the those receiving placebo, participants who received intranasal diversity of OT therapeutic effects by reviewing relevant OT prior to initial viewing of face images of Black and White animal and human background studies and summarizing individuals subsequently exhibited no race difference in the very promising results of intranasal OT compared to accurate identification of previously seen faces. The placebo treatment in RCTs conducted in patients with schi- Blandon-Gitlan et al. report seems to conflict with the zophrenia and alcohol dependence. OT significantly reduced implication of De Dreu's studies that the prosocial effects of psychotic symptoms in patients with schizophrenia in all intranasal OT are exclusively in-group directed and empha- three published studies in which intranasal treatments were sizes the importance of understanding and comparing the administered twice daily for 2, 3 or 8 weeks. Furthermore, in underlying psychobiology in situations in which OT amplifies one study (Pedersen et al., 2011) OT improved performance in or diminishes group differences. some social cognitive domains and, in another study (Feifel The generally prosocial effects of OT in animal and human et al., 2012), OT improved performance in one neurocognitive studies has generated considerable interest in examining OT domain, verbal learning. The latter findings are particularly as an etiological factor in and a potential treatment for exciting because social cognitive and neurocognitive deficits developmental and psychiatric disorders in which social contribute substantially to social dysfunction which is a deficits are prominent. Anagnostou et al. (this volume) report major cause of disability in schizophrenia (Fett et al., 2011) the results of a small (15 subjects) preliminary 12-week and does not respond to currently available randomized, placebo-controlled of intranasal (Bellack et al., 2004; Penn et al., 2009). Pedersen OT in high functioning children and adolescents with autism presents his recent preliminary evidence that intranasal OT spectrum disorders (ASDs). A number of measures of social is remarkably effective in blocking withdrawal in highly cognition and functioning as well as repetitive behaviors and alcohol-dependent patients. He presents the rationale for anxiety improved in OT recipients over the treatment period. testing intranasal OT efficacy in decreasing drinking in Gains in some measures persisted 3 months after disconti- alcohol dependence. Pedersen poses and discusses numerous nuation of intranasal OT treatment. No adverse events, questions about our remarkable lack of information about OT OT-related side effects or clinically significant changes in systems in the and heretofore unexamined OT laboratory tests occurred. In a recently published 6-week mechanisms that may be important in the pathophysiology randomized, placebo-controlled pilot clinical trial in 19 adults of psychotic disorders and addictions. Finally, Pedersen with ASDs, Anagnostou et al. (2012) found that subjects speculates that antipsychotic and withdrawal-blocking effi- receiving twice daily intranasal OT compared to placebo cacy of OT may be linked to as yet unexplored OT mechan- demonstrated improvements in empathic accuracy, lower isms that were selected during the evolution of maternal and order repetitive behaviors and quality of life. Francis et al. other social behaviors in placental mammals. (this issue) review the remarkable evidence that, in develop- MacDonald and Feifel (this issue) explore the broad topic of mental disorders with markedly different underlying genetic OT as a pathophysiological factor as well as a potential abnormalities and pathophysiology, OT function is signifi- treatment of anxiety disorders. These issues are particularly cantly altered in ways that correspond to the social behavior important because disabling levels of anxiety occur in many anomalies exhibited by individuals with those disorders. psychiatric illnesses. The authors begin with a discussion of Specifically, the marked social avoidance and anxiety in the vague and inconsistent ways in which anxiety is referred ASDs, the Prader–Willi Syndrome and the Fragile-X Syndrome to and conceptualized across disciplines followed by a review are associated with impaired OT (and in some cases AVP) of preclinical studies of OT and anxiety. An evolution-based activity and/or release in patients or animal models of the theoretical framework is then developed about the centrality disorders. In contrast, in individuals with Williams Syndrome of OT and other socially relevant neuropeptides and neuro- who exhibit hypersociability, a positive correlation is found transmitters in regulating the balance between fear, anxiety, between OT levels and increased stranger approach and avoidance and aggression-generating brain systems that have decreased adaptive social behavior. Francis et al. hypothesize been selected to enhance self-preservation and , Editorial / brain research 1580 (2014) 1–7 5

approach, affiliation-generating systems that have been uncharted territory that hopefully will inspire new directions selected to increase social/reproductive success. MacDonald in human and clinically-relevant animal research on OT and and Feifel continue with a highly comprehensive review of psychopathology. To this end, many of the authors point out studies of intranasal OT effects on as well as the relationship the enormous gaps in our current understanding of OT of OT levels and OT receptor gene variants with anxiety systems in the human brain, propose novel hypotheses about measures, stress hormone levels and anxiety-generating brain the mechanisms of OT psychotherapeutic effects and discuss region activity in normal subjects as well as individuals with the possible etiology of those effects in heretofore unexplored anxiety disorders or anxiety phenotypes. This includes the roles for which OT may have been selected during the authors' trend level findings that intranasal OT reduced evolution of placental mammalian maternal and other social anxiety in male subjects (but exacerbated anxiety in female behaviors. As the Editors, we hope the papers appearing in subjects) in a small (N 13), preliminary 3-week RCT in this special issue will collectively generate new and fruitful ¼ patients with generalized anxiety disorder. While acknowl- discussions toward future research aimed at understanding edging that connections between brain regions play important and treating social deficits in psychopathological conditions. roles in anxiety, the potential significance of OT activity in several brain areas in anxiety and anxiety disorders is thor- references oughly examined. The authors leave the reader with ten as yet unanswered questions to guide future work on OT and anxiety. Overall, a strong case is made to conduct clinical Anagnostou, E., Soorya, L., Chaplin, W., Bartz, J., Halpern, D., trials testing the efficacy of intranasal OT in anxiety disorders. Wasserman, S., Wang, A.T., Pepa, L., Tanel, N., Kushki, A., Soeken et al. (this issue) make the case that dysfunction of Hollander, E., 2012. Intranasal oxytocin versus placebo in the OT systems may contribute to and treatment of adults with autism spectrum disorders: a randomized controlled trial. Mol. Autism 3 (1), 16. that OT may be an effective treatment in this disorder. The Andari, E., Duhamel, J.R., Zalla, T., Herbrecht, E., Leboyer, M., authors summarize current treatment options and discuss Sirigu, A., 2010. Promoting social behavior with oxytocin in their limitations. Animal and human research is reviewed high-functioning autism spectrum disorders. Proc. Natl. Acad. supporting the efficacy of OT and the poten- Sci. USA 107, 4389–4394. tial role of OT in depression especially in the postpartum Atzil, S., Hendler, T., Zagoory-Sharon, O., Winetraub, Y., Feldman, period. Areas requiring further research are identified. R., 2012. Synchrony and specificity in the maternal and the paternal brain: relations to oxytocin and vasopressin. J. Am. There is broad consensus among evolution theorists that Acad. Child Adolesc. Psychiatry 51, 798–811. higher primate and human intelligence is the product of Barrett, L., Henzi, P., 2005. The social nature of primate cognition. selection for mental abilities that provide advantages in Proc. R. Soc. B 272, 1865–1875. assessing complex and variable social situations and adjust- Bartz, J.A., Zaki, J., Bolger, N., Ochsner, K.N., 2011. Social effects of ing social behavior to enhance access to resources and oxytocin in humans: context and person matter. Trends Cogn. mating opportunities (Barrett and Henzi, 2005; Moll and Sci. 15 (7), 301–309. Tomasello, 2007; Pinker, 2010; Heyes, 2012) in often highly Baumgartner, T., Heinrichs, M., Vonlanthen, A., Fischbacher, U., Fehr, E., 2008. Oxytocin shapes the neural circuitry of trust fluid social milieus. The primary function of the human brain and trust adaptation in humans. Neuron 58, 639–650. and its advanced mental abilities, such as abstract thought Bellack, A.S., Schooler, N.R., Marder, S.R., Kane, J.M., Brown, C.H., and language, is viewed as enabling social success. The social Yang, Y., 2004. Do clozapine and risperidone affect social brain concept of human intelligence thus far has had mini- competence and problem solving?. Am. J. Psychiatry 161, mal impact in biological psychiatry and even less influence 364–367. on CNS development by the pharmaceutical industry Boccia, M.L., Goursand, A.-P.S., Bachevalier, J., Anderson, K.D., (Macdonald and Feifel, 2013; but see also Brüne, 2008). The Pedersen, C.A., 2007. Peripherally-administered non-peptide oxytocin antagonist, L368,899s,accumulatesinlimbicbrain extensive evidence cited in this and other articles in this areas: a new pharmacological tool for the study of social Special Issue suggest that CNS OT systems are a key compo- motivation in non-human primates. Hormon. Behav. 52, 344–351. nent of the human social brain. The rapidly accumulating Boccia, M.L., Petrusz, P., Suzuki, K., Marson, L., Pedersen, C.A., evidence that central OT abnormalities are found in many 2013. Immunohistochemical localization of oxytocin receptors psychiatric and developmental disorders and that OT treat- in human brain. Neuroscience 253, 155–164. ment reduces the severity of symptoms in a wide range of Born, J., Lange, T., Kern, W., McGregor, G.P., Bickel, U., Fehm, H.L., disorders supports the view that malfunctions of the social 2002. Sniffing neuropeptides: a transnasal approach to the human brain. Nat. Neurosci. 5, 514–516. brain contribute to many psychopathophysiological condi- Bosch, O.J., Meddle, S.L., Beiderbeck, D.I., Douglas, A.J., Neumann, tions (Brüne, 2008). Intranasal OT administration and other I.D., 2005. Brain oxytocin correlates with maternal aggression: methods of examining brain OT systems may provide unique link to anxiety. J. Neurosci. 25, 6807–6815. opportunities to test the significance of this central element Bosch, O.J., Neumann, I.D., 2008. Brain vasopressin is an of the social brain in psychiatric disorders. The high degree of important regulator of maternal behavior independent of tolerability and safety of intranasal OT administration adds dams’ trait anxiety. Proc. Natl. Acad. Sci. USA 105, to its feasibility and appeal as a means of probing the normal 17139–17144. Bru¨ ne, M., 2008. Textbook of Evolutionary Psychiatry: The Origins and psychiatrically-disabled human social brain. of Psychopathology. Oxford University Press, Oxford. At the request of the Editors, articles in this special Chang, S.W.C., Barter, J.W., Ebitz, R.B., Watson, K.K., Platt, M.L., volume have gone beyond reviewing and critiquing studies 2012. Inhaled oxytocin amplifies both vicarious reinforcement on OT in human social behavior and psychopathology by and self reinforcement in rhesus macaques (Macaca mulatta). articulating novel and innovative concepts and illuminating Proc. Natl. Acad. Sci. 109, 959–964. 6 Editorial / brain research 1580 (2014) 1–7

Champagne, F.A., 2012. Interplay between social experiences and increases maternal aggressive behavior. Behav. Neurosci. 117, the genome: epigenetic consequences for behavior. Adv. 195–201. Genet. 77, 33–57. Macdonald, K., Feifel, D., 2013. Helping oxytocin deliver: Consiglio, A.R., Borsoi, A., Pereira, G.A., Lucion, A.B., 2005. Effects considerations in the development of oxytocin-based of oxytocin microinjected into the central amygdaloid nucleus therapeutics for brain disorders. Front. Neurosci. 7, 35. and bed nucleus of stria terminalis on maternal aggressive McCarthy, M.M., 1990. Oxytocin inhibits infanticide in female behavior in rats. Physiol. Behav. 85, 354–362. house mice (Mus domesticus). Hormon. Behav. 24, Consiglio, A.R., Lucion, A.B., 1996. Lesion of hypothalamic 365–375. paraventricular nucleus and maternal aggressive behavior in McCarthy, M.M., Bare, J.E., vom Saal, F.S.V., 1986. Infanticide and female rats. Physiol. Behav. 59, 591–596. parental behavior in wild female house mice: effects of Domes, G., Heinrichs, M., Michel, A., Berger, C., Herpertz, S.C., ovariectomy, adrenalectomy and administration of oxytocin 2007. Oxytocin improves “mind-reading” in humans. Biol. and prostaglandin F2α. Physiol. Behav. 36, 17–23. Psychiatry 61, 731–733. McGowan, P.O., Szyf, M., 2010. Environmental epigenomics: Ebitz, R.B., Watson, K.K., Platt, M.L., 2013. Oxytocin blunts social understanding the effects of parental care on the epigenome. vigilance in the rhesus macaque. Proc. Natl. Acad. Sci. 110, Essays Biochem. 48 (1), 275–287. 11630–11635. Meyer-Lindenberg, A., Domes, G., Kirsch, P., Heinrichs, M., 2011. Feifel, D., MacDonald, K., Cobb, P., Minassian, A., 2012. Adjunctive Oxytocin and vasopressin in the human brain: social intranasal oxytocin improves verbal memory in people with neuropeptides for translational medicine. Nat. Rev. Neurosci. schizophrenia. Schizophr. Res. 139, 207–210. 12 (9), 524–538. Ferguson, J.N., Young, L.J., Hearn, E.F., Matzuk, M.M., Insel, T.R., Mikolajczak, M., Pinon, N., Lane, A., de Timary, P., Luminet, O., Winslow, J.T., 2000. Social amnesia in mice lacking the 2010. Oxytocin not only increases trust when money is at oxytocin gene. Nat. Genet. 25, 284–288. stake, but also when confidential information is in the Fett, A.K., Gracia-Dominguez, M., Viechtbauer, W., Penn, D.L., van balance. Biol. Psychol. 85, 182–184. Os, J., Krabbendam, L., 2011. A meta-analytic study on Moll, H., Tomasello, M., 2007. Cooperation and human cognition, social cognition, and functional outcome in non- intelligence: the Vygotskian intelligence hypothesis. Philos. affective psychosis. Neurosci. Biobehav. Rev. 35, Trans. R. Soc. B 362, 639–648. 573–588. Neumann, I.D., Maloumby, R., Beiderbeck, D.I., Lukas, M., Figueira, R.J., Peabody, M.F., Lonstein, J.S., 2008. Oxytocin receptor Landgraf, R., 2013. Increased brain and plasma oxytocin after activity in the ventrocaudal periaqueductal gray modulates nasal and peripheral administration in rats and mice. anxiety-related behavior in postpartum rats. Behav. Neurosci. Psychoneuroendocrinology 38 (10), 1985–1993. 122, 618–628. Neumann, I.D., Torner, L., Wigger, A., 2000. Brain oxytocin: Fischer-Shofty, M., Shamay-Tsoory, S.G., Harari, H., Levkovitz, Y., differential inhibition of neuroendocrine stress responses and 2010. The effect of intranasal administration of oxytocin on anxiety-related behaviour in virgin, pregnant and lactating fear recognition. Neuropsychologia 48, 179–184. rats. Neuroscience 95, 567–575. Giovenardi, M., Padoin, M.J., Cadore, L.P., Lucion, A.B., 1998. Parr, L.A., Modi, M., Siebert, E., Young, L.J., 2013. Intranasal Hypothalamic paraventricular nucleus modulates maternal oxytocin selectively attenuates rhesus monkeys’ attention to aggression in rats: effects of ibotenic acid lesion and oxytocin negative facial expressions. Psychoneuroendocrinology 38, antisense. Physiol. Behav. 63, 351–359. 1748–1756. Guastella, A.J., Einfeld, S.L., Gray, K.M., Rinehart, N.J., Tonge, B.J., Pedersen, C.A., 2013. Oxytocin regulation of maternal behavior: Lambert, T.J., Hickie, I.B., 2010. Intranasal oxytocin improves from rodents to humans. In: Choleris, E., Pfaff, D., Kavaliers, emotion recognition for youth with autism spectrum M. (Eds.), Oxytocin, Vasopressin and Related Peptides in the disorders. Biol. Psychiatry 67, 692–694. Regulation of Behavior. Cambridge University Press, Guastella, A.J., Mitchell, P.B., Dadds, M.R., 2008. Oxytocin Cambridge, pp. 148–182. increases gaze to the eye region of human faces. Biol. Pedersen, C.A., Caldwell, J.D., Walker, C., Ayers, G., Mason, G.A., Psychiatry 63, 3–5. 1994. Oxytocin activates the postpartum onset of rat maternal Heyes, C., 2012. New thinking: the evolution of human cognition. behavior in the ventral tegmental and medial preoptic areas. Philos. Trans. R. Soc. B 367, 2091–2096. Behavioral Neuroscience 108, 1163–1171. Holman, S.D., Goy, R.W., 1995. Experiential and hormonal Pedersen, C.A., Gibson, C.M., Rau, S.W., Salimi, K., Smedley, K.L., correlates of caregiving in rhesus macaques. In: Pryce, C.R., Casey, R.L., Leserman, J., Jarskog, L.F., Penn, D.L., 2011. Martin, R.D., Skuse, D. (Eds.), Motherhood in Human and Intranasal oxytocin reduces psychotic symptoms and Nonhuman Primates: Biosocial Determinants 3rd Schultz– improves theory of mind and social perception in Biegert Symposium, Kartause Ittingen, Switzerland, schizophrenia. Schizophr. Res. 132, 50–53. September 1994. Karger, Basel Swizterland, pp. 87–93. Penn, D.L., Keefe, R.S., Davis, S.M., Meyer, P.S., Perkins, D.O., Hurlemann, R., Patin, A., Onur, O.A., Cohen, M.X., Baumgartner, Losardo, D., Lieberman, J.A., 2009. The effects of antipsychotic T., Metzler, S., Dziobek, I., Gallinat, J., Wagner, M., medications on emotion perception in patients with chronic Maier, W., Kendrick, K.M., 2010. Oxytocin enhances schizophrenia in the CATIE trial. Schizophr. Res. 115, amygdala-dependent, socially reinforced learning and 17–23. emotional empathy in humans. J. Neurosci. 30, 4999–5007. Petrovic, P., Kalisch, R., Singer, T., Dolan, R.J., 2008. Oxytocin Kendrick, K.M., 2000. Oxytocin, motherhood and bonding. Exp. attenuates affective evaluations of conditioned faces and Physiol. 85, 111S–124S. amygdala activity. J. Neurosci. 28, 6607–6615. Kendrick, K.M., Da Costa, A.P.C., Broad, K.D., et al., 1997. Neural Pinker, S., 2010. The cognitive niche: coevolution of intelligence, control of maternal behaviour and olfactory recognition of sociality, and language. Proc. Natl. Acad. Sci. USA 107 (Suppl. offspring. Brain Res. Bull. 44, 383–395. 2), S8993–S8999. Kosfeld, M., Heinrichs, M., Zak, P.J., Fischbacher, U., Fehr, E., 2005. Riem, Madelon M.E., Bakermans-Kranenburg, Marian J., Pieper, Oxytocin increases trust in humans. Nature 435, 673–676. Suzanne, Tops, Mattie, Boksem, Maarten A.S., Vermeiren, Lubin, D.A., Elliott, J.C., Black, M.C., Johns, J.M., 2003. An oxytocin Robert R.J.M., van IJzendoorn, Marinus H., Rombouts, Serge A. antagonist infused into the central nucleus of the amygdala R.B., 2011. Oxytocin modulates amygdala, insula, and inferior Editorial / brain research 1580 (2014) 1–7 7

frontal responses to infant crying: a randomized Cort A. Pedersenn controlled trial. Biol. Psychiatry 70, 291–297. Department of Psychiatry, University of North Carolina at Chapel Riem, M.M.E., Van IJzendoorn, M.H., Tops, M., Boksem, M.A., Hill, CB 7160, Chapel Hill, NC 27599-7160, USA Rombouts, S.A., Bakermans-Kranenburg, M.J., 2012. No E-mail address: [email protected] laughing matter: intranasal oxytocin administration changes functional brain connectivity during exposure to infant laughter. Neuropsychopharmacology 37, 1257–1266. Steven W.C. Chang Smith, A.S., A˚ gmo, A., Birnie, A.K., French, J.A., 2010. Center for Cognitive Neuroscience, Levine Science Research Center, Manipulation of the oxytocin system alters social behavior and attraction in pair-bonding primates, Callithrix penicillata. Duke University, Durham, NC, USA Horm. Behav. 57, 255–262. Takayanagi, Y., Yoshida, M., Bielsky, I.F., et al., 2005. Pervasive social deficits, but normal parturition, in oxytocin receptor- Christina L. Williams deficient mice. Proc. Natl. Acad. Sci. USA 102, 16096–16101. Department of Psychological & Brain Sciences, Theodoridou, A., Rowe, A.C., Penton-Voak, I.S., Rogers, P.J., 2009. Duke University, Durham, NC, USA Oxytocin and social perception: oxytocin increases perceived facial trustworthiness and attractiveness. Horm. Behav. 56, 128–132. Winslow, J.T., Insel, T.R., 1991. Social status in pairs of male 0006-8993/$ - see front matter squirrel monkeys determines the behavioral response to & 2014 Published by Elsevier B.V. central oxytocin administration. J. Neurosci. 11 (7), 2032–2038. Winslow, J.T., Noble, P.L., Lyons, C.K., Sterk, S.M., Insel, T.R., 2003. Rearing effects on cerebrospinal fluid oxytocin concentration http://dx.doi.org/10.1016/j.brainres.2014.07.033 and social buffering in rhesus monkeys. Neuropsychopharmacology 28, 910–918. Young, L.J., Wang, Z., 2004. The neurobiology of pair bonding. Nat. Neurosci. 7 (10), 1048–1054.

nCorresponding author. brain research 1580 (2014) 78–101

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Approaching the biology of human parental attachment: Brain imaging, oxytocin and coordinated assessments of mothers and fathers

J.E. Swaina,b,c,n, P. Kimd, J. Spicere, S.S. Hoa, C.J. Daytona,f, A. Elmadihg, K.M. Abelg aDepartment of Psychiatry, University of Michigan, USA bCenter for Human Growth and Development, University of Michigan, USA cDepartment of Psychology, University of Michigan, USA dDepartment of Psychology, University of Denver, USA eDepartment of Psychiatry, Columbia University, USA fSchool of Social Work, Wayne State University, USA gCentre for Women's Mental Health, Institute of Brain Behaviour and Mental Health, Manchester Academic Health Sciences Centre, University of Manchester, UK article info abstract

Article history: Brain networks that govern parental response to infant signals have been studied with Accepted 7 March 2014 imaging techniques over the last 15 years. The complex interaction of thoughts and Available online 15 March 2014 behaviors required for sensitive parenting enables the formation of each individual's first social bonds and critically shapes development. This review concentrates on magnetic Keywords: resonance imaging experiments which directly examine the brain systems involved in Parent–child relationship parental responses to infant cues. First, we introduce themes in the literature on parental Brain imaging brain circuits studied to date. Next, we present a thorough chronological review of state-of- fMRI the-art fMRI studies that probe the parental brain with a range of baby audio Attachment and visual stimuli. We also highlight the putative role of oxytocin and effects of Parenting psychopathology, as well as the most recent work on the paternal brain. Taken together, Caregiving a new model emerges in which we propose that cortico-limbic networks interact to support Oxytocin parental brain responses to infants. These include circuitry for arousal/salience/motivation/ Maternal reward, reflexive/instrumental caring, emotion response/regulation and integrative/complex Paternal cognitive processing. and the quality of caregiving behavior are likely determined by the responsiveness of these circuits during early parent-infant experiences. The function of these circuits is modifiable by current and early-life experiences, hormonal and other factors. Severe deviation from the range of normal function in these systems is particularly associated with (maternal) mental illnesses – commonly, depression and

Abbreviations: MPOA, medial ; BNST, bed nucleus of the stria terminalis; dlPFC, dorsolateral ; IPV-PTSD, interpersonal violence-related posttraumatic stress disorder; nACC, ; ACC, anterior cingulate; IFG, inferior frontal gyrus; NAcc, nucleus accumbens; OFC, orbitofrontal cortex nCorrespondence to: Rachel Upjohn Building, 4250 Plymouth Road, Ann Arbor, MI 48109-2700, USA. E-mail address: [email protected] (J.E. Swain). http://dx.doi.org/10.1016/j.brainres.2014.03.007 0006-8993/& 2014 Elsevier B.V. All rights reserved. brain research 1580 (2014) 78–101 79

anxiety, but also schizophrenia and bipolar disorder. Finally, we discuss the limits and extent to which brain imaging may broaden our understanding of the parental brain given our current model. Developments in the understanding of the parental brain may have profound implications for long-term outcomes in families across risk, resilience and possible interventions. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2014 Elsevier B.V. All rights reserved.

1. Introduction to the parental brain of variation between individual mothers, despite being rela- tively stable in the same mothers across time and contexts Human mothers and fathers exhibit a repertoire of parental (Behrens et al., 2012; Jaffari-Bimmel et al., 2006; Wan et al., thoughts and activities that manifest interesting similarities 2013). Maternal sensitivity represents a pattern of behavior and differences as part of an array the crosses cultures (Hrdy, which provides the infant with its primary social experience. 2000) and species across species (Clutton-Brock, 1991) with This suggests that it is important for organizing and regulat- the common goal of caring for offspring. For humans, besides ing the infant's emotional, social and cognitive systems and meeting the primal evolutionary needs for survival and is consistent with accumulating evidence that maternal continuation of our species, parenting involves interrelated sensitivity predicts a range of social-emotional child out- biological, psychological, and behavioral caregiving mechan- comes – including the quality of their attachment relation- isms that contribute critically to the first environment the ships (Bakermans-Kranenburg et al., 2003; De Wolff and van child experiences as a new family member. Ijzendoorn, 1997), self-regulation (Eisenberg et al., 2001), The rich animal literature on parental brain (Bridges, 2008) social functioning (Kochanska, 2002; Van Zeijl et al., 2006), strongly supports the notion that a coherent understanding socio-emotional development (De Wolff and van Ijzendoorn, of the physiology governing parenting in humans is also 1997), and cognitive and language competence (Bernier possible. Indeed, much of the research on the human par- et al., 2010; Tamis-LeMonda et al., 2001). Furthermore, the ental brain involves attempts to look for human homologs of absence of skills needed to respond sensitively to child 'parental' brain circuits found in animals. This approach has signals has been linked to risk for maltreatment (Milner, initially been successful using audiovisual stimuli of babies 1993, 2003). Poor maternal sensitivity in infancy predicts during functional imaging – particularly with contrasts of later harsh parenting (Joosen et al., 2012) and attitudes own vs. unknown babies – to activate brain regions that towards punishment (Engfer and Gavranidou, 1987). recognize and respond to the special salience for the parent Frightening and anomalous maternal behavior confers pro- of their own baby's stimuli (Barrett and Fleming, 2011; Swain, found risk to the parent–infant attachment relationship 2011b; Swain et al., 2007). (Schuengel et al., 1999) to subsequent child outcomes and In this review, we summarize the state-of-the art of human future parenting of their own children. parental brain imaging in mothers and fathers. More specifi- The concept of parent–infant attachment represents a land- cally, we focus on the interpretation of brain-responses to baby- mark of contemporary developmental psychology (Bowlby, cry and baby visual stimuli respectively and consider attempts 1969b, 1973). In fact, Bowlby formulated his attachment theory to integrate findings within the cognitive, affective, and social after studying associations between maternal deprivation and neurosciences. First, we selectively review the psychological juvenile delinquency, postulating a universal human need to and imaging evidence of relevant executive, emotion response/ form close, affect-laden bonds, primarily between mother and regulation and reward mechanisms that may be activated in infant. He strongly argued, from an evolutionary perspective, the service of parental sensitivity. We then discuss these that attachment represents an innate biological system promot- fi findings in the context of research on oxytocin, and review ing proximity-seeking between an infant and a con-speci c fi recent attempts to understand how parenting difficulties or attachment gure. This proximity then increases the likelihood even psychopathology may be understood as abnormalities or of survival to reproductive age. malfunctions of parental brain circuits. The importance of this Because of this powerful biological instinct for attachment, fi mechanistic understanding of parenting is underscored by and in response to the patterns of attachment identi ed in the consideration of effects on offspring and possibilities to identify Ainsworth mother–infant studies (Ainsworth et al., 1978), families at risk and optimize therapeutic interventions. Bowlby (1977) hypothesized that all human infants attach to their caregiver but that children manifest different patterns of attachment “security” depending on the quality of the care 1.1. Parental sensitivity and child attachment they receive. Indeed, a vast literature in the study of attach- ment over last several decades has established that infants of Ainsworth et al. (1978) first defined maternal sensitivity as a caregivers who are available, responsive and sensitive to their mother's ability to attend and respond to her child in ways emotional and physical needs tend to manifest patterns of that are contingent to the infant's needs. In the naturalistic “secure attachment.” Conversely, chaotic, unpredictable, context, sensitive maternal care behaviors show a great deal rejecting or neglectful care in which non-contingent responses 80 brain research 1580 (2014) 78–101

to the child occur frequently, result in insecure or disorganized parenting – an issue yet to be studied. Attention bias to patterns of attachment evolves (Shaver et al., 1987). infant distress has also been compared between breastfeed- Understanding the neurobiology of attachment–formation ing (n 27) and formula feeding (n 24) mothers of 3–6 ¼ ¼ through parental sensitivity may help to formulate and month old infants and observed to be greater in breastfeeding ameliorate pervasive and complex social problems such as mothers (Pearson et al., 2011). It is plausible that reduced child abuse and neglect. However, little is known about the attention bias towards emotional stimuli (specifically infant cognitive or neurobiological mechanisms which underpin distress) is associated with low maternal sensitivity in some healthy sensitivity, let alone poor sensitivity in mothers with mothers. However, other mothers with low sensitivity may mental illness. This may explain why promising interven- show a selectively exaggerated attention bias to infant dis- tions targeted at improving parenting through improved tress which results in the mother becoming overwhelmed by sensitivity have shown inconsistent findings and generally the stimuli; this may be particularly relevant for mothers and small effect sizes (Bakermans-Kranenburg et al., 2003; Wan fathers with schizophrenia or in those with little social et al., 2008c). supports in deprived circumstances. In depressed patients, evidence suggests a tendency to pay more attention to 1.2. Neurocognitive mechanisms underlying parental negative emotional stimuli (Elliott et al., 2011); an effect sensitivity mediated by enhanced response in the ventral anterior cingu- late, which may contribute to the maintenance of low mood. For sensitive caregiver responses to infant cues, complex brain systems must manage an array of complex thoughts 1.2.2. Emotion regulation and behaviors contingent on feedback from the baby. These Recognizing emotion in preverbal infants is more difficult than include recognition and acknowledgment of child signals, recognizing emotions in adults. In some parents, inability attribution of salience to child cues, maintenance of visual to recognize and distinguish the subtleties of infant emotion contact, expression of positive effect, appropriate mirroring cues may underpin poor maternal sensitivity. Consistent with and vocal quality, resourcefulness in handling child's distress this, depression is associated with decreased discrimination of or expanding the interaction, consistency of style, and dis- facial emotion (Anderson et al., 2011). Response to distressing play of an affective range that matches the infant's readiness signals from the infant also requires a mother to distinguish to interact. Such behaviors are likely to be the result of positive from negative emotions; indeed, studies suggest that a complex neural networks involved in generating and organiz- mother's sensitivity to distress may be a better predictor of ing emotional responses (Kober et al., 2008) as well as in child outcomes than her sensitivity to non-distress cues attention and executive function, reward and motivation, (Joosen et al., 2012; Leerkes, 2011; Leerkes et al., 2009; and sensorimotor circuits – a model well supported by the McElwain and Booth-Laforce, 2006). Thus, poor maternal care literature (Swain and Lorberbaum, 2008). We update this behavior could derive in part from reduced recognition, as well model below incorporating the newest work on fathers and as reduced response to infant emotions generally, and/or the effects of mental illness on parenting capacity. Indeed, specifically, to signals of infant distress. Conversely, some mental health problems may manifest as impairment in any mothers may become overwhelmed by their infant's distress. of these networks and play a significant role in poor parent- Notably, enhanced responsiveness to negative emotions, ing. Identifying these may provide an opportunity to predict (mediated by enhanced amygdala response), has been which parents are likely to have which type of parenting observed in non-parent depression (Arnone et al., 2012). In problems and facilitate targeting of interventions. Here, we depressed mothers, studies suggest women may avoid or limit highlight brain processes, mechanistically related to certain exposure to distressing infant stimuli (Field, 2010; Murray et al., cortico-limbic circuits, and relevant for healthy parenting 1996; Pearson et al., 2012). In anxiety and depression, modulat- sensitivity. ing stress and emotional responsiveness is an important target for treatment and is associated with clinical improvement 1.2.1. Executive function (Harmer et al., 2011). The importance of emotion regulation Executive functions, including attention control, working in the responses to baby stimuli is also consistent with the memory and flexible task-switching, are likely to be impor- ideas of postpartum preoccupations discussed below (Section tant to parental sensitivity. Deficits in attention set-shifting, 1.2.4). spatial working memory and a sustained attention measure have been linked with poor maternal sensitivity to non- 1.2.3. Reward/motivation distress infant cues (Gonzalez et al., 2012). In closely related Extensive recent review of the animal literature (Numan work, mothers with a classification of disorganized attach- and Woodside, 2010) suggests that response to infants forms ment responded more slowly to negative attachment words a model motivational system employing and and the speed of response to such stimuli was correlated with oxytocin-rich pathways such as the medial preoptic area disorganization, suggesting negative associations with (MPOA). Through such pathways, infant cues are thought to attachment stimuli that may contribute to ongoing cognitive provide motivation for maternal care behavior. Reward pro- difficulties during mother–infant interactions (Atkinson et al., cesses include immediate hedonic responses (“liking”) and 2009). Greater attention bias to infant distress cues in late approach motivation (“wanting”) or learning (Berridge and has also been associated with better scores on a Kringelbach, 2008). Frontostriatal brain regions are also criti- parental bonding questionnaire (Pearson et al., 2010), raising cally implicated in reward, in particular the orbitofrontal the question of how attention deficit disorder affects cortex (OFC) (Rolls, 2004) and ventral striatum including brain research 1580 (2014) 78–101 81

nucleus accumbens (NAcc) (Born et al., 2011). Although the developing emotional bonds with the infant, and risk inade- OFC generally codes hedonic signals, the medial OFC is quate parenting (Feldman et al. 1999, 2009). Maternal schizo- particularly important for computing reward value while phrenia also interferes with care giving behavior in ways that the lateral OFC makes stronger contributions to reward merit separate study and treatment (Abel et al., 2005; Wan learning. In mothers, the initial experience of pleasure and et al., 2008a, 2008b, 2008c). For typically developing parents, activity in these brain circuits when exposed to their own the story for fathers will likely differ from mother. For infant's cues may increase the salience of their infant's example, unlike for mothers, paternal sensitivity was pre- stimuli and promote greater attention and bond-formation dicted by anxious as well as caregiving and positive thoughts to ensure continuous engagement in sensitive caregiving. (Kim et al., 2013). More nuanced assessments of mother and Indeed, reward/motivation pathways have been shown to be father thought, mood and behaviors are required to clarify active in response to baby-stimuli (discussed below). In how they are related over time and how they might inform mothers with low sensitivity, the motivational or incentive the nascent study of paternal brain physiology (below). salience of emotional and/or infant cues may be diminished Parental empathy, (appropriate perception, experience and through deficits in this and also in other OT- reward response to another's emotion) may be especially relevant for pathways (Curley, 2011). Thus, in rhesus monkeys, mothers preverbal infants. Deficits of empathy in disorders such as with greater attachment to infants possess the G allele of the schizophrenia and autism have been associated with abnorm- OPRM1 m-opioid receptor and have higher oxytocin responses alities in certain brain networks (Gallese et al., 2004; Iacoboni, to lactation and pregnancy (Higham et al., 2011). Strikingly, 2009; Uddin et al., 2007). These systems overlap significantly activity of at the m-opioid receptor is also central to with brain responses of parents to infant stimuli reviewed the processing of reward in the context of drug dependence here i.e. cingulate and insular cortices (Bernhardt and Singer, (Herz, 1997; Simmons and Self, 2009), and social behaviors 2012). (Higham et al., 2011). Very few studies, however, have exam- Experiencing pain personally and experiencing the pain of a ined neural substrates of reward-related processes among loved one activate insula and anterior cingulate, with mothers who display disrupted caregiving behavior. another's pain activating more anterior regions and own pain also recruiting brainstem, cerebellum, and sensorimotor cor- 1.2.4. Parental thoughts – preoccupations/habits, empathy tex. Such decoupled, yet parallel representations of empathy and positive thoughts in cortical structures like insula and anterior cingulate are Parents experience dynamic change in their thoughts and postulated as necessary for our ability to mentalize, that is, behaviors oriented toward their new infants. Immediately understand the thoughts, beliefs, and intentions of others in after a child's birth and during the first few months of the relation to ourselves (Frith and Frith, 2003; Hein and Singer, infant's life, parents are particularly drawn to their infants' 2008). Humans may utilize separate circuitry to “decouple” vocalizations and physical attributes and focus their thoughts representations of external vs. internal information in order to on the infants' physical and psychological needs (Bowlby, understand physical properties and assess personal emotional 1969a; Winnicott, 1956). Such intense mental focus and values. In support of this, a brain network consistently behaviors were first referred to as “primary maternal preoccupa- activated during tasks that require mentalizing has emerged, tion” by Winnicott (1956) and seem also consistent with the including dorsomedial prefrontal cortex (DMPFC), medial pre- activation of motivational-reward pathways by infant cues frontal cortex (MPFC), precuneus/posterior cingulate cortex reflecting an important evolutionary mechanism through (PCC), temporoparietal junction (TPJ), and posterior superior which infants promote long-term emotional ties with their temporal sulcus (pSTS) (Frith and Frith, 2006; Mitchell, 2009). parents (Atkinson et al., 2009; Feldman et al., 1999; Gonzalez Such a framework for understanding of the contribution of et al., 2012; Leckman and Mayes, 1999). social ties to broad health issues (Eisenberger and Cole, 2012) Leckman et al. (1999) reported that all domains of parental promises to be very helpful in considering the importance of preoccupation peaked right after childbirth and then began to parent-infant dyads in health promotion. decrease over the course of the first three to four postpartum By three to four months postpartum, infants are more months in mothers and fathers, albeit less intensely for socially interactive, and parents increasingly engage in recipro- fathers. Since newborns have very limited ways of commu- cal positive interactions. These positive interactions further help nicating their needs, parents must constantly check the mothers to strengthen their attachment and heighten the infants in order to identify and attend to those needs. experience of positive feelings toward their infants (Mercer, The intensity of parental preoccupations decreases as par- 1985). As they successfully feed, take care of, and build affec- ents gain more experience in parenting and infants become tionate connections with their infants, mothers develop positive more responsive. However, excessively and persistently high feelings and self-confidence about parenting (Benedek, 1954). levels of parental preoccupation, particularly anxious and Fathers may develop attachment to their infants more gradually intrusive thoughts and harm-avoidant behaviors can be (Anderson, 1996; Pruett, 1998). Positive feelings about their associated with postpartum psychopathology such as post- infants and parenting experience maybe critical in pathways partum obsessive compulsive disorders (OCD). In one sample, engaging dopamine–oxytocin reward circuits described in ani- levels of parental preoccupations were related to postpartum mal models (MacDonald, 1992; Numan and Insel, 2003b; anxiety at the first month postpartum (Feldman et al., 1999) Shahrokh et al., 2010). Thus, interactions with the infant may and lower maternal sensitivity at 3–4 months postpartum enhance parental oxytocin and dopamine release and foster the (Kim et al., 2013). Abnormally low levels of parental preoccu- maintenance of positive parental behaviors with associated pation, e.g. in postpartum depression also pose difficulties in attentiveness and sensitive caregiving – perhaps implicating 82 brain research 1580 (2014) 78–101

future therapeutic potential (Macdonald et al., 2013). In contrast, capture more complex brain responses, but also average if interactions with the infant are negative and stressful, parents them out making subtle responses more difficult to detect. may experience less brain activation in reward-motivation Brain activity during these blocks may then be measured and pathways, find the relationship less satisfying, harbor fewer compared between periods of attending to stimuli of interest positive parental thoughts and be less willing to maintain it. and control stimuli to generate maps of the brain indicating Such reward pathways maybe relevant very early in the differences in brain activity that may be important for one set postpartum, as mother's positive feelings towards her unborn of perceptions and thoughts vs. another. With parents as fetus as well as her perception of her fetus have been subjects, infant cries and pictures have been stimuli and associated with greater maternal sensitivity to the infant's comparisons of brain activity measured during baby cry vs. signals and more affectionate vocalizations and touch (Keller control sound experience have been said to relate to the et al., 2003; Keren et al., 2003). It may be interesting in the parental experience of a baby cry, and so the associated future to examine prepartum thoughts as relevant for post- parenting thoughts and behaviors. Many of these studies partum behaviors. Idealization of the infant and positive make use of seed-based analyses with modest group sizes thoughts about parenting may positively reinforce both that will require replication in different populations to permit quality of parenting and feelings of reward. On the other generalization of findings and all results could relate to hand, worrying about the infant – such as health and other activity that is excitatory or inhibitory. Future imaging with concerns – may be associated with lower parental sensitivity connectivity analyses also promise to define more than just and higher intrusiveness. Indeed, negative feelings about response regions, but circuit behavior involving coordinated parenting may be linked to the parents' difficulties in devel- activity between distant brain areas (Sripada et al., 2013). oping emotional bonds with their infants (Mercer, 1985; The important limitation that response to baby cues may Nystrom and Ohrling, 2004) although the nature and direction not be directly related to parental care is being addressed of this link needs clarification. in studies that correlate responses to parental care The following sections review the current state of evidence described below. from experiments designed to elucidate the brain basis of parental attachment by presenting infant stimuli – sometimes emotionally charged – during brain imaging. Such studies have 2. Parental brain responses to infant stimuli incorporated psychological, behavioral and hormonal mea- sures and a very few include parents with abnormalities in 2.1. What's in a baby-cry? parenting or mental illness. The power and sensitivity of functional imaging deliver mechanistic understanding of In addition to communicating basic levels of discomfort, abnormalities in brain associated with poor parenting and hunger, and pain (Soltis, 2004), baby cry also signals the need promise earlier detection of problems that may not be obvious for physical proximity of caregivers leading to the dynami- from behavioral or self-report measures. In the future, brain- cally interactive cry-care thoughts and behaviors that lead to based understandings may allow refinement of tailored treat- attachment (Swain et al., 2004b), for which new parents ment approaches toward parenting and targeted augmenta- appear to be primed (Bowlby, 1969b). Supportive studies have tion of parenting resiliency (Rutter, 2013; Swain et al., 2012). demonstrated that human mothers can recognize the cries of Tables 1 and 2 summarize experiments to date on human their own infants, and mothers and fathers activate brain in parents using baby sound and visual stimuli with brain fMRI. areas hypothesized to be involved in mammalian parenting behavior (Swain and Lorberbaum, 2008; Swain et al., 2007). 1.3. Human brain imaging methods 2.1.1. Brain responses to auditory baby-stimuli: The primary technique used to study the brain in this review baby-cry – 1st 10 years is non-invasive blood-oxygen-dependent functional magnetic Building on the thalamocingulate theory of maternal beha- resonance imaging (fMRI). fMRI assays brain activity by vior in animals (MacLean, 1990), Lorberbaum et al. (1999) - in indirectly measuring changes in regional blood oxygenation the first, though rather small study of its kind - predicted and during different periods in which infant cues may be pre- found that baby cries selectively activate cingulate and sented. The differences between a region's oxygenated and in mothers in the early postpartum months deoxygenated hemoglobin, between states of action vs. inac- exposed to an audio taped 30 s standard baby cry using fMRI. tion for instance, provide characteristic magnetic signals Informed by the animal literature, they expanded their localized to millimeters that are detected by scanners posi- hypotheses to include the basal forebrain's medial preoptic tioned around each subject's head. An important caveat area and ventral bed nucleus of the stria terminalis and its throughout the interpretation of parenting fMRI studies, rich reciprocal connections as being critical to parental however, is that that brain activity measurements represent behaviors (Lorberbaum et al., 2002). These include the des- an integration of electrical brain activity that may be instan- cending connections which modulate more basic reflexive taneous yet the related blood flow change lags behind over caring behaviors such as nursing, licking, grooming and seconds. Furthermore, experimental design captures brain carrying reflexes in rodent studies, and ascending connec- activity over periods of a few seconds or 10's of seconds. tions such as the mesolimbic and mesocortical dopamine On the one hand, short blocks or events may capture briefly systems involved in more general motivation and flexible held mental states, but miss bigger changes such as sus- responses required to tend a crying infant or prepare for a tained emotion; while on the other hand longer blocks may threat. Even though these first studies involved cry stimuli brain research 1580 (2014) 78–101 83

Table 1 – Human parent brain responses to infant stimuli.

Author, year Parent Age of Paradigm, variable Study Brain network group N infants design Arousal Reflex Emotion Cognition Salience Care Regulation

Baby-cry, Mothers Lorberbaum 4 3 weeks to Other vs. control 30 s YYY Y et al. (1999) 3.5 years Blocks Lorberbaum 10 1–2 months Other vs. control 30 s YYY Y et al. (2002) Blocks Seifritz et al. 10 o3 years Other vs. control, 6s YNY Y (2003) ve/ ve Events þ À Swain et al. 11–14 2–4 weeks Own vs. other, 30 s YYY Y (2003, 2004a, and 3–4 experience and Blocks 2006) months thoughts Swain et al. 12 2–4 weeks Own vs. other, delivery 30 s YYY Y (2008) Blocks Kim et al. 26 2–4 weeks Own vs. control, early- 30 s YNY Y (2010a) life Blocks Kim et al. 20 2–4 weeks Own vs. other, feeding 30 s YYY Y (2011) Blocks Laurent 22 18 months Own vs. other, HPA axis 21 s NYY Y et al. (2011) Blocks Venuti et al. 9 (of 18) 44years Hunger cry vs. noise, 10 s NNY Y (2012) atypical cry from Events infants with autism Baby cry, fathers Seifritz et al. 10 o3 years Other vs. control, 6s YYY Y (2003) ve/ ve Events þ À Swain et al. 92–4 weeks Own vs. other, 21 s YYY Y (2003, 2004a) and 3–4 experience and Blocks months thoughts Mascaro 36 1 or 2 years Other baby-cry vs. 5s Y Y Y Y et al. (2013a) control sound events De Pisapia 9M 9F 1 year Hunger cry; males vs. N N N Y þ et al. (2013) females 14 s Blocks Baby cry, maternal- pathology Laurent and 22 18 months Own vs. other, 21 s YYY Y Ablow (2012) postpartum depression Blocks Musser et al. 22 18 months Own vs. other, maternal 21 s YYY Y (2012) sensitivity Blocks Landi et al. 54 2 months Other vs. noise, high 2s YYY Y (2011) and low distress cry, Events substance abuse

Parental brain papers are comprehensively listed in chronological order and according to responses mothers, fathers and psychopathology – to baby cry with columns for parent group, time point (age of infant), experimental paradigm, and study design. Brain networks related to parenting featured in the article are also indicated. Empty boxes indicate lack of significance in the networks defined as (i) Arousal/Salience amygdala, ventral striatum, ¼ (ii) Reflex Care , ¼ (iii) Emotion Regulation mPFC, ACC, and ¼ (iv) Cognition dlPFC, insula, inferior frontal and orbitofrontal gyri, temporoparietal junction. ¼

Glossary for Table: activations and deactivations, measured by functional magnetic resonance imaging, satisfied significance criteria of random effects analysis at po0.05 or fixed effects analysis at po0.001 at a minimum.

that did not originate from the parent's own infant, and the behavior in animals (Numan and Insel, 2003a) and opened control sounds were emotionally negative (sounded like up the field. harsh static on the television), significant brain responses fit Hypothesizing that parental gender and experience would with existing knowledge about regulation of parenting also influence neural responses to baby sounds such as baby 84 brain research 1580 (2014) 78–101

Table 2 – Human parent brain responses to infant stimuli.

Author, year Parent Age of Paradigm, variable Study Brain network group N infants design Arousal Reflex Emotion Cognition Salience Care Regulation

Mothers Baby Visuals Swain et al. (2003, 9–14 2–4 weeks Own vs. other, 30 s Blocks Y Y Y Y 2004a, 2004b, 2005, and 3–4 experience and under review months thoughts Bartels and Zeki (2004) 19 9 months to Own vs. other, 15 s Blocks Y Y Y Y 6 years comparison with romantic partner Leibenluft et al. (2004) 75–12 years Own vs. other 1.5 s Event Y Y Y Y Ranote et al. (2004) 10 4–8 months Own vs. other 20–40 s Y Y N N Nitschke et al. (2004) 62–4 months Own vs. other, affect 30 s Blocks N N Y Y Strathearn et al. (2005, 28–30 3–18 months Own vs. other, 2 s Events Y Y Y Y 2008, 2009) affect, OT Noriuchi et al. (2008) 13 15–20 Own vs. other, 32 s Video Y Y Y Y months distressed Lenzi et al. (2009) 16 6–12 months Own vs. other, joy/ 2 s Events Y Y Y Y distress Atzil et al. (2011) 28 4–6 months Own vs. other, 2 min YYY Y parenting syn Video Barrett et al. (2012) 22 3 months Own vs. other, 3 s Events Y Y Y Y parenting Riem et al. (2011, 2012) 21 (Non- N/A Cry (2 days) vs. 10 s Events Y Y Y Y mom) control Atzil et al. (2012) 15 4–6 months Own videos, 2 min YYY Y synchrony, OT Video Lahey et al. (2012) 35 4–6 years Own vs. other, 13 s Video N N Y Y parenting Fathers Baby visuals Swain et al. (2003, 9–14 2–4 weeks Own vs. other, 30 s Blocks Y Y Y Y 2004a, 2004b, 2005, and 3–4 experience and under review) months thoughts Atzil et al. (2012) 15 4–6 months Own videos, 2 min YYY Y synchrony, OT Video Kuo et al. (2012) 10 2–4 months Own vs. other 15 s Video Y Y Y Y videos, Baby vs. Doll videos Mascaro et al. (2013b) 70 1–2 years Own vs. other 14 s Blocks, YYY Y images, measure T, regional father-behav Maternal Baby psychopathology visuals Moses-Kolko et al. 30 12 weeks Emotion faces, 4 s Blocks Y Y Y Y (2010) depression Schechter et al. (2012) 20 12–42 Own4other, IPV- 40 s Video Y Y Y Y months PTSD Moser et al. (2013) 20 12–42 Own4other, IPV- 40 s Video Y Y Y Y months PTSD Laurent and Ablow 22 18 months Own vs. other – 18 s Blocks Y Y Y Y (2013) depression

Parental brain papers are comprehensively listed in chronological order and according to responses mothers, fathers and psychopathology – to visual stimuli with columns for parent group, time point (age of infant), experimental paradigm, and study design. Brain networks related to parenting featured in the article are also indicated. Empty boxes indicate lack of significance in the networks defined as (i) Arousal/Salience amygdala, ventral striatum, ¼ (ii) Reflex Care hypothalamus, ¼ (iii) Emotion Regulation mPFC, ACC, and ¼ (iv) Cognition dlPFC, insula, inferior frontal and orbitofrontal gyri, temporoparietal junction. ¼

Glossary for Table: activations and deactivations, measured by functional magnetic resonance imaging, satisfied significance criteria of random effects analysis at po0.05 or fixed effects analysis at po0.001 at a minimum. brain research 1580 (2014) 78–101 85

cry and laughter, Seifritz et al. (2003) studied four groups: included midbrain, basal ganglia, cingulate, amygdala and mothers and fathers of children under age 3, and non-parent insula (Swain et al., 2003). This may reflect an increase in males and females, with 10 subjects in each group. They used arousal, obsessive/anxiety circuits normally more active for an event-related fMRI design, measuring brain responses to new parents and persistently sensitive in some mental ill- brief 6-s sounds. Over the entire sample, intensity-matched nesses (Swain et al., 2007). Interview and self-report studies baby sounds of crying and laughing compared to “neutral” have reported that mothers are significantly more preoccu- sounds (white noise pulsed at 5-Hz with an averaged fre- pied than fathers, and this is consistent with relatively quency spectrum similar to the infant vocalizations) pro- greater activation of amygdala and basal ganglia reported in duced more brain activity in bilateral temporal regions. These mothers compared with fathers (Swain et al., 2004a, under regions might be important for hearing processes (Heshyl's review). In addition, grouping mothers and fathers together gyrus and temporal poles), processing human vocalizations, across experience and comparing activity at 2–4 weeks vs. 3–4 and empathic emotion processing including emotional mem- months postpartum for own baby cry showed that significant ory. They also reported that women as a group including, activity shifted from the amygdala and insula to medial parents and non-parents (but not men), showed decreased prefrontal cortical and hypothalamic (including hormonal activity in response to both baby cry and laughter in the control) regions. This fits well with changes in parenting subgenual anterior cingulate cortex. This finding is contrary confidence/experience in healthy parents, as a mother learns to the other studies (Lorberbaum et al., 1999, 2002; Swain et al., to associate her infant cries with more flexible social beha- 2005; Swain and Lorberbaum, 2008) and highlights how sample viors and more mature attachment, there is greater regula- selection, choice of stimuli and precision of region of interest tory cortical brain activity and less alarm and anxiety-related examined might affect findings. It is also the case that a 6 s vs. activity (amygdala and insula). a 30 s stimulus time may have different meanings to new parents or that there may be non-linear or multiphasic 2.1.2. Brain responses to auditory baby-stimuli: responses within anterior cingulate similar to the well- baby-cry – last 5 years recognized phasic responses seen in amygdala. Their within- Consistent with accumulating evidence of the importance of group analyses reported that parents activated more to infant oxytocin (OT) in establishing and regulating parenting in crying than laughing in the right amygdala, while non-parent humans (Galbally et al., 2011), neuroimaging in humans response was greater for infant laughing than crying (Seifritz suggests brain circuits that respond to baby-stimuli are also et al., 2003). Although there was no direct comparison between associated with OT pathways. In one study, mothers experi- parents and non-parents, these within-group results suggest encing vaginal vs. cesarean delivery – as a proxy for higher vs. that being a parent might be associated with changes lower OT – showed greater brain activity in response to own in amygdala function and represents the first attempts to vs. other baby-cry at 2–4 weeks postpartum in emotion include gender and experience-dependent aspects of human regulation and limbic regions, including the caudate, thala- parenting. mus, hypothalamus, amygdala and pons (Swain et al., 2008). Taking this approach further, Swain et al. (2003) have This fits with evidence for cesarean section being associated been gathering data on groups of new parents across a range with increased risk of postpartum depression – a condition of of experience, temperament and parent–infant interaction aberrant emotion processing (Groenewold et al., 2012). At 3–4 styles using each parent's own baby cries and including months postpartum, the same mothers – all of whom comprehensive interviews and self-reports. In this design, remained healthy – no longer showed differing responses to parents underwent brain fMRI during 30 s blocks of infant own baby-cry (Swain, 2011a), suggesting that any potential cries generated by their own infant contrasted with a “stan- early-postpartum abnormalities in oxytocin are reversible. dard” cry and control noises matched for pattern and inten- However, the relationship between Cesarean section and PPD sity. In addition, they added a longitudinal component with is strongly confounded by a range of factors including low scans and interviews at 2 time points: 2–4 weeks and 12–16 maternal age, past depression and low social supports. weeks postpartum to coincide with the transition to parent- Recent studies have also used breastfeeding as a proxy for hood associated with peaks of parental preoccupation in the maternal OT levels Imaging studies using baby cry (Kim et al., early postpartum (Leckman et al., 1999). They hypothesized 2011), find that breastfeeding vs. formula feeding is asso- that parental responses to own baby cries would include ciated with greater activations to own baby cries vs. other specific activations in thalamo-cortico-basal ganglia circuits baby cries in anterior and posterior cingulate, thalamus, believed to be involved in human ritualistic and obsessive– midbrain, hypothalamus, septal regions, dorsal and ventral compulsive thoughts and behaviors (Baxter, 2003; Leckman striatum, medial prefrontal cortex, right orbitofrontal/insula/ et al., 2004). They also reasoned that emotional alarm, temporal polar cortex region, and right lateral temporal arousal and salience detection centers including amygdala, cortex and fusiform gyrus. Additionally, when cry response and insula (Britton et al., 2006; LeDoux, 2003) was compared with the inter-stimulus rest periods, instead of would be particularly activated by own baby cry. The experi- the control sound (which some mothers judged to be aver- mental block design was used in order to give parents a sive), the amygdala was active. Furthermore, and for the first chance to reflect on their experience of parenting and, time, Kim et al. (2011) reported a correlation of brain activity according to our hypothesis, become more preoccupied with to own vs. unknown baby-cry with independently-rated their infants' well-being and safety. In a group of first-time behavioral measures of parenting in response (amygdala) mothers (n 9) at 2–4 weeks postpartum, regions that were and regulation (frontal cortex) regions, suggesting the impor- ¼ relatively more active with own vs. unknown contrast tance of balanced responses for sensitive care giving. Such 86 brain research 1580 (2014) 78–101

studies have led to the suggestion that, combined with process the sound information than men. However, such behavioral strategies, acute or prolonged OT treatment may gender difference was not detected in response to cries of offer a safe and accessible intervention for women at risk of ASD infants – perhaps because of abnormalities in ASD-cries postpartum depression although no randomized controlled that are easier to detect. Alternatively, high distress or atypical trials have emerged. cries may inhibit maternal brain responses in some women Among nulliparous women, acute administration of oxy- particularly if they are rendered more sensitive to negative tocin has been reported to decrease amygdala responses and valence stimuli, such as with depression (Groenewold et al., increase insula and inferior frontal gyrus responses to poten- 2012). tially adverse stimuli i.e. infant cries (Riem et al., 2011). Indeed, in the first study of baby cry in women who Similarly, acute was also found to increase experienced a postpartum depression, mothers at 18 months insula response to infant cries (Bos et al., 2010) in nulliparous postpartum showed reduced responses to own vs. other baby healthy women. Testosterone is metabolized to , in regions that process reward stimuli and promote maternal which activates neural regions important for maternal moti- motivation – nucleus accumbens, caudate and thalamus vation. Therefore, increasing oxytocin and testosterone avail- (Laurent and Ablow, 2012). More depressive symptoms were ability in maternal brain may modulate the processing of also associated with less activity in response to own vs. other distressing emotional information e.g. infant cries and facil- baby cry in the OFC, dorsal ACC, and superior frontal gyrus – itate more appropriate responses, especially in mothers with regions important for emotion information processing and dysregulated responses to such stimuli (see above). By con- regulation, including valence and salience. Similarly, mothers trast, greater reactivity to stress has also been who used one or more teratogenic substances during preg- associated with reduced neural responses to own baby cries nancy (e.g. tobacco and alcohol) showed decreased responses (vs. other baby cries) among primiparous mothers (Laurent to low distress cries in prefrontal cortex, insula and amygdala et al., 2011). The reduced neural responses were detected in (Landi et al., 2011). These reports may imply that maternal the regions important for emotion regulation (anterior cingu- depression and addiction to substances are associated with late cortex and medial PFC) and maternal motivation (limbic disrupted mother-infant relationships and reduced neural area and periaqueductal gray). Studies such as these impli- activation to infants, including infant cries, although whether citly address effects of early-life events on later parental or how such associations are causal remains to be identified. brain function, translating to humans some aspects of Alternatively, abnormalities in reward pathways that con- well-established rodent and non-human primate models tributed to the likelihood of substance use might also affect (Champagne, 2010; Kaffman and Meaney, 2007; Veenema, parenting independently of these substances. In any case, 2009) as well as long-term effects of early trauma on stress reduced activations in the PFC may considered a candidate axis functioning in human studies (Lupien et al., 2009; biomarker associated for risk of harsh, neglectful or other McEwen, 2008). Thus, in mothers who report higher maternal forms of negative parenting among high risk women such as care in their own experience of childhood, greater responses mothers with mental illness (Goodman et al., 2011; Molitor to infant cries may be seen in regulatory cortical regions, and Mayes, 2010) and represent a future potential therapeutic including emotion regulation areas of the middle and super- target. ior frontal gyri, whereas mothers reporting lower perceived Parenting styles in humans are likely to be transmitted maternal care may be more likely to show increased hippo- across generations (Belsky et al., 2005; Champagne, 2010) campal activations (Kim et al., 2010b) This underlines the which explains in part why the neural and behavioral varia- importance for parenting of coordinated responses of these tions of mothering are associated with a mothers' own cortical and subcortical regions, as well as the importance of parental care experiences. Quality of own parenting therefore early life experience in the functioning of these circuits. acts as a specific early life environment that should perhaps be Recent work has also examined whether variations in targeted for interventions for primiparous and multiparous infant cries are associated with maternal neural response to mothers who report low quality of maternal care. Such new the cries. One study exposed nulliparous women to two mothers exhibit both reduced brain density and activation to different (high and low) distress levels of infant cries infant cries in frontal, orbital and temporal cortices compared (Montoya et al., 2012). Interestingly, women showed greater to mothers who reported high quality maternal care (Kim neural responses to low distress cries vs. high distress cries in et al., 2010b). Preliminary evidence also suggests that during the superior and middle temporal gyri. The findings suggest the first few critical months postpartum, the same regions that, compared to high distress cries, it was harder for increase in size among healthy new mothers (Kim et al., 2010a) women to interpret the possible causes and meaning of the according to voxel based morphometry. low distress cries. Another study exposed a group of men and Parenting also influences the development of attachment women (a half of the total participants were parents) to cries style in children, which in turn affects the development of a of typically developing infants and cries of infants who later child's social information processing. For example, insecurely were diagnosed with autism spectrum disorder (ASD) (Venuti attached nulliparous women displayed greater amygdala et al., 2012). Cries of the ASD children showed abnormal reactivity to infant cry sounds than securely attached counter- features including high frequencies, which elicited more parts (Riem et al., 2012), suggesting that infant cries represent negative feelings in healthy adults and women, not men, more aversive stimuli for them. Amygdala responses to infant showed greater deactivations in dorsal medial PFC and cry in new mothers have also been interpreted as a form of posterior cingulate cortex in response to cries of healthy heightened sensitivity (Barrett and Fleming, 2011). It seems infants, indicating that women recruited more attention to that amygdala responses may be associated with negative and brain research 1580 (2014) 78–101 87

positive valence brain responses – perhaps relating to unmea- maternal brain responses of own vs. familiar child to contrasts sured factors in maternal experience. of best friend vs. familiar friend to control for familiarity and positive affect and argued that responses were unique to the 2.2. What's in a baby's face? own child's stimuli. They proposed that parent–infant attach- ment was regulated by a “push–pull” mechanism involving Not only is facial recognition of one's infant and their selective activation of motivation and reward systems, with emotional state critical to their survival, but evidence sup- cortical regions suppressing critical social assessment and ports human preference for the exaggerated cute/infantile negative emotion systems (Bartels and Zeki, 2004); this they features of baby faces which appear to activate reward argued may be extended to orchestrate positive feelings and mechanisms within the caregiver and motivate/promote caring behaviors. mothers to caregiving, bonding and attachment. In addition In another report, photograph stimuli of much older own to activation of nucleus accumbens (NAcc) in response to vs. other children (5–12 years old) were used as stimuli and baby schema (Glocker et al., 2009) and medial orbitofrontal mothers were asked to focus on identity but not feelings cortex (OFC) response to infant faces relative to adults during scanning (Leibenluft et al., 2004). Some social cogni- (Kringelbach et al., 2008), higher levels of infantile features tion regions which are important for empathy were signifi- are preferred (Parsons et al., 2011). Neuroimaging also sug- cantly activated in this paradigm, including anterior gests that the brain responds differentially to infant, relative paracingulate, posterior cingulate and superior temporal to adult human and adult and infant animal faces, in non- sulcus (Saxe, 2006). In another small fMRI experiment using parents, in regions including lateral premotor cortex, supple- visuals across familiarity, Nitschke et al. (2004) studied six mentary motor area, cingulate cortex, anterior insula and healthy, primiparous mothers' at 2–4 months postpartum as thalamus (Caria et al., 2012). Interestingly, activation in OFC they viewed smiling pictures of their own and unfamiliar and fusiform face area is disrupted when an alteration to the infants. They reported OFC activations that correlated posi- structure of an infant face (i.e. cleft lip) is perceived (Parsons tively with pleasant mood ratings to infant pictures. In et al., 2013). Once parenthood is established, data further contrast, areas of visual cortex that also discriminated indicate that greater specialization of the brain occurs with between own and unfamiliar infants were unrelated to mood for example a right-sided lateralization in the prefrontal ratings (Nitschke et al., 2004). Perhaps, activity in the OFC – cortex (PFC) for emotional discrimination of infant relative which may vary across individuals – is involved with high to adult faces (Nishitani et al., 2011), perhaps as a result of order dimensions of maternal attachment. The implication attention processing (Thompson-Booth et al., 2014). that complex aspects of parenting could be quantified using fMRI of frontal areas to predict risks of mood problems in 2.2.1. Brain responses to visual baby stimuli – 1st 10 years parents is appealing but requires further detailed work. Most early fMRI studies of parental responsiveness have very Ranote et al. (2004) conducted a similar small experiment small samples and detail few characteristics of mothers or with the innovative, and perhaps more ecological video fathers making findings less reliable than some larger, newer (silent) infant stimuli in 10 healthy mothers viewing alter- studies. We shall outline the older literature first before nating 40 s blocks of own neutral, and an unknown infant. considering more rigorous recent papers. One set of studies They reported significant activation in “own” vs. “unknown” used photographs taken extremely early (i.e. 0–2 weeks post- infant comparison in the left amygdala and temporal pole partum), by the parents themselves. Using a block design, and interpreted involvement of circuitry regulating emotion mothers and fathers 2–4 weeks postpartum saw 6 pictures and theory-of-mind (ability to attribute feelings and states of continuously for 5 s each for blocks of 30 s (Swain et al., 2003, mind to others). This fits with fMRI experiments on biological 2006) and own vs. other baby picture contrasts revealed motion, which activate similar temporal cortex regions activations in frontal and thalamo-cortical circuits. Correla- (Morris et al., 2005). tions between activations to own vs. other contrast and Considering the contribution of the infant's affect to parent–infant interactions also revealed significant activations maternal brain function prompted a study, again using silent in superior temporal lobe, OFC and ventral tegmental areas. video clips of own vs. other infants in play or separation These networks may be important for regulation of parental situations (Noriuchi et al., 2008). First, these authors con- motivation and reward associated with baby-directed empa- firmed increased activation associated with own baby pic- thy, approach and caring behaviors, as well as social bonding. tures, in cortical orbitofrontal, anterior insula and precuneus To examine whether parental love may make use of the areas, as well as subcortical regions including periaqueductal same reward circuits as other forms of love, Bartels and Zeki gray and putamen. These areas activate in arousal and (2004) used photographs of own, familiar and unfamiliar infants reward learning. Second, they found strong and specific (9 months to 3.5 years of age) as stimuli for parent brains. In 20 differential responses of mother's brain to her own-infant's healthy mothers viewing still face photographs of their own distress in , caudate nucleus, thalamus, child compared to age-matched photographs of other children, posterior and , anterior cingulate, they reported increased activity in midbrain periaqueductal dorsal regions of OFC, right inferior frontal gyrus and dor- gray and substantia nigra regions, dorsal and ventral striatum, somedial prefrontal cortex – regions involved in emotion thalamus, left insula, ORC, sub-, pre-, and supra-genual anterior regulation and habitual behavioral response systems that cingulate, and superior medial prefrontal cortex. There were are active in a range of normal and abnormal emotion- also increases in cerebellum, left fusiform, and left occipital control states including obsessive–compulsive disorder cortex, but decreases in the left amygdala. They also compared (Swain et al., 2007). They also found correlations in OFC with 88 brain research 1580 (2014) 78–101

own baby-response and happiness as well as to their own mothers with secure attachment showed greater activation of distressed baby response in superior temporal regions. This is brain reward regions, including ventral striatum, and consistent with the emerging importance of these areas in oxytocin-rich hypothalamic/pituitary regions. These results social behaviors. Socially salient, personally tailored images chime with effects of own parenting on fMRI activations to and video-clips are now being combined with behavioral infant stimuli and suggesting that individual differences in measures to understand better the functional architecture maternal attachment experiences might also be crucial to of parental brain. Taken together, fMRI experiments with measure in future imaging studies, and perhaps that they are parents, especially those using own vs. other/unknown baby linked with development and integration of dopaminergic visual stimuli commonly activate emotion/motivation- and oxytocinergic neuroendocrine systems in striatum and reward areas along with cortical regulation areas, thus laying hypothalamus. the groundwork to test hypotheses about how such brain In last few years, parental brain circuitry has been studied structures regulate parental thoughts and behaviors. in the context of variables associated with parental illness For example, empathy may be conceived as a key parent- (such as depression, anxiety, and substance misuse), and the ing thought process and studied in mothers observing and work highlights regions-of-interest such as the amygdala, imitating faces of their own and someone else's child (Lenzi anterior cingulate cortex (ACC), PFC, insula and striatum. In a et al., 2009). In this study, regions believed to contain mirror comparison of own-positive vs. unfamiliar-positive infant neurons and connected limbic regions, including insula and images, left amygdala response was reduced as a function amygdala respectively, were more active during emotional of poorer concurrent maternal experience (measured by expressions from own child. Furthermore, the insula depression, anxiety, parental distress and attachment- response correlated with the empathy-related measure of related feelings about the infant) (Barrett et al., 2012). In a maternal reflective function. They also reported that baby joy study employing similar task conditions, reduced dorsal (d) expressions across identity evoked mostly right limbic and ACC activation to own-distress images was reported in paralimbic areas important to emotional processing, whereas depressed relative to non-depressed primiparous women (as ambiguous expressions elicited responses in left-sided, high measured by the Structured Clinical Interview for DSM-IV) order cognitive and motor areas, logically reflecting asso- (Laurent and Ablow, 2013). As a function of increased depres- ciated cognitive effort and preparation to respond. sive symptomatology Center for Epidemiological Studies In addition to the capacity forcomplexempathicthoughts Depression (CESD) measure, reduced response was observed toward their infant, parents require motivation to undertake in the OFC and insula to own-joy faces and in left PFC and interactive behaviors and to derive reward from interacting insula/striatum to own-joy vs. own-distress faces. Though with their infants. Integrating brain areas relevant for these infant stimuli were not included, brain response to negative actions with key hormones is likely to reinforce behaviors and emotional faces (fear and anger) in depressed vs. healthy provide for optimal parental sensitivity under a range of postpartum women has been examined, yielding observa- circumstances. Considering some of these complex factors, 28 tions such as reduced dorsomedial (dm)PFC activation in healthy, first-time, singleton mother–infant dyads at 5–10 depressed vs. healthy women to faces; a negative correlation months postpartum were involved in one series of studies between left amgydala and depression severity in depressed using visual infant facial cues of varying affect (smiling, neutral women; a positive correlation between right amygdala and and crying) and scanning at 7–17 months postpartum. Notably, absence of infant-related hostility in depressed women; and this is well past the 3monthpostpartumthresholdfor negative functional connectivity between left dmPFC and left $ sophisticated social dyadic interactions (Strathearn et al., amygdala in healthy, but not depressed women (Moses-Kolko 2005, 2008). Dopamine-associated reward-processing regions et al., 2010). Finally, in a study examining substance users in were activated when mothers viewed their own vs. an the postpartum period relative to non-substance users, unknown infant's face, including ventral tegmental area/sub- reduced activation to infant faces was reported in dorsolat- stantia nigra, and striatum. In addition, there were eral (dl)PFC, ventrolateral (vl)PFC, occipital regions, parahip- responses in emotion processing (medial prefrontal, anterior pocampus and amygdala (Landi et al., 2011). These studies cingulate, and insula cortex), cognition (dorsolateral prefrontal suggest that maternal adversity and/or substance misuse cortex), and motor/behavioral outputs (primary motor area). may specifically be associated with reduced activation in Furthermore, happy, but not neutral or sad own-infant faces, reward and emotion circuits, with some indication that top- activated nigrostriatal brain regions interconnected by dopami- down connectivity between the PFC and limbic regions also nergic neurons, including substantia nigra and dorsal putamen. may be compromised (Moses-Kolko et al., in press). Finally, a region-of-interest analysis revealed that activation in these regions was related to positive infant affect (happy4neu- 2.2.3. Brain responses to mother–baby dyad movies tral4sad) for each own-unknown infant-face contrast. Recent studies have employed mother–child dyadic vignettes as visual stimuli in the study of the parental brain. In one 2.2.2. Brain responses to visual baby stimuli – last 5 years small but intriguing study, parental adversity, as defined by In pursuit of the role of oxytocin in modulating the parental interpersonal violence-related posttraumatic stress disorder brain, Strathearn et al. (2009) studied a total of 30 first-time (IPV-PTSD), was examined. Motivated by findings demonstrat- mothers using a variety of affect-laden baby picture stimuli in ing atypical caregiving in PTSD mothers following separation, combination with the Adult Attachment Interview and per- vignettes of own vs. unfamiliar toddlers during play and ipheral plasma oxytocin responses to infant play. In response following separation were presented to healthy mothers and to their own infant's smiling and crying faces during fMRI, those with IPV-PTSD (Schechter et al., 2012). IPV-PTSD mothers brain research 1580 (2014) 78–101 89

vs. healthy mothers (11 vs. 9) reported greater stress to viewing inconsistencies between studies mean many questions separation vignettes, greater limbic and reduced fronto- remain to be answered. First, we need better understanding cortical activity was observed in IPV-PTSD mothers relative of how the brain manages the presentation of static indivi- to healthy mothers in separation vs. play conditions, with dual baby images relative to moving images involving the reports of stress to viewing separation linking to the neural dyad. Does different responses to a vignette reflect something findings. These results are consistent with those in depression specific to the participant or to the vignette itself? Second, we indicating reduced fronto-cortical activity in response to own have to understand how findings from neuroimaging studies vs. other baby-cry (Laurent and Ablow, 2012). A follow-up examining perception of infant stimuli relate to the capacity study of the same subjects reported correlation between limbic for affective cognition of individual mothers and fathers. brain activations in response to own-child video vignettes and Future neuroimaging studies assessing parental affective symptoms (Moser et al., 2013), suggesting disso- cognition alongside behavioral measures of parenting may ciative symptoms as a mechanism for reduced maternal change our view of a dedicated parental circuitry. The find- sensitivity among mothers affected by IPV. More work with ings from extant parental brain studies that have employed mothers affected by high-risk circumstances, such as IPV, reverse inference in interpretation would be strengthened by depression and anxiety is needed to clarify and substantiate studies that select more constrained and better described findings and begin to inform treatments. samples using more sophisticated paradigms. Third, imaging So far, however, a few other studies have assessed paradigms have explored the effects of only a few forms of parental brain response to mother–infant dyadic vignettes parental adversity or mental illness. It is difficult to design a as a function of optimal vs. non-optimal parenting. Atzil et al. study which can selectively examine or isolate the effects of, (2011) categorized participants with a synchronous, relative for example, the severe chronic stress of poverty, young to intrusive, parenting style and asked them to view vignettes maternal age and early life trauma, and no studies have yet of their own infants; of themselves interacting with their focused on resilience or the rescuing of adversity following infants and those of unfamiliar infants and dyads. In another intervention. We still know far too little about how or study, when own vs. other infant vignettes were compared, whether behavioral or neural correlates of maternal sensitiv- maternal synchrony was associated with significantly ity/maternal responsiveness are modifiable. However we increased activity of left NAcc, while maternal intrusiveness have reason to be cautiously optimistic given recent brain was linked to activity of right amygdala (Atzil et al., 2012). imaging studies of parents and the development of biomar- However, in this experiment, it is unclear whether this kers for depression and its treatment (Harmer et al., 2011). pattern of brain response was a function of synchronous Among possible medical treatments for low maternal sensi- mothers or the viewing of a synchronous mother–infant tivity, oxytocin has been proposed. The following section interaction. This group has also examined the role of oxyto- examines points of intersection between the parental brain cin, reporting positive correlations between cross-sectional and oxytocin, and possible mechanisms that might inform plasma oxytocin levels and activation in the left NAcc, left studies which examine modulation of maternal responding insula, left intraparietal lobule, left and right temporal cor- in brain and behavioral paradigms through the use of tices, left sgACC, and left NAcc and right amygdala in oxytocin. mothers, and a range of frontal cortical areas in fathers. More studies with rigorous identification of maternal caregiving quality and parenting style with brain imaging are required 3. Parenting – connections between brain and (Wan et al., 2014). oxytocin (OT) Accumulating research confirms what most people, and nearly all mothers know to be the case: infant stimuli occupy The posterior pituitary neuropeptide hormone, oxytocin (OT), a privileged status for the adult human brain, which may be is a well-recognized component of a complex biobehavioral modeled as in Fig. 1. Auditory and visual signals from system crucial for the emergence of mother–infant bonding preverbal infants can enhance brain regions important to (Ross and Young, 2009). Research in rodents and other recognition, emotion and motivation (i), reflexive caring mammals has highlighted the importance of OT (on a back- behaviors (ii), emotion regulation (iii), and empathy/theory ground of changing levels of and of mind toward sensitive caregiving and attachment in order during pregnancy and labor) to facilitate the onset and to increasing the likelihood that the vulnerable infant will maintenance of maternal behavior (Champagne et al., 2001, survive into adulthood and successfully navigate the social 2003; Champagne, 2008; Insel and Young, 2001; Rosenblatt world and find a mate. and Ceus, 1998). This was first suggested from studies that The early study of parental brain employed a variety of reported display of “full maternal behavior” in virgin female baby audiovisual stimuli to elucidate a putative neurocircui- rats injected with OT (Pedersen and Prange, 1979). Conver- try for parenting. More recently, studies have added correla- sely, inhibition of postpartum maternal behavior was tions with psychometrics of parenting, parental adversity affected in rats by injecting them with an OT-receptor (e.g. poor maternal experience, attachment, perinatal depres- antagonist (van Leengoed et al., 1987). Among high “licking sion and PTSD) as well as indices of parenting quality (e.g. and grooming” (i.e. maternal caregiving) female dams, sig- synchronicity and maternal sensitivity) and finally neuroen- nificantly higher levels of OT receptors were also seen in docrine measures (oxytocin). brain regions implicated in the expression of maternal In spite of this encouraging scenario, important problems behavior across species, during pregnancy, at parturition with study design, gaps in the evidence-base and and when nursing pups, such as the central nucleus of the 90 brain research 1580 (2014) 78–101

BABY FACTORS A. INFANT “INPUT”. STIMULI: Cry, Visuals, Touch, Smell

B. CORTICO-LIMBIC MODULES

Prematurity Early Life Delivery Experience Experience -parenting Breastfeeding -education -socioeconomic status

Psycho- therapy Current Life Experience -social support -marital circumstances -mental health Anti - vs. psycho- pathology

C. PARENTAL ATTACHMENT “OUTPUT” BEHAVIORS -corticolimbic modules interact and may be active simultaneously -flexible cortical regions such as prefrontal and sensorimotor cortices – likely more important in humans than non-humans –orient, regulate affect and generate parent-infant interpersonal synchrony and sensitive parenting thoughts and behaviors

Fig. 1 – Human parental circuits. Brain regions expected to be important to human parenting. This is based on human and animal studies. Based on brain imaging of parents to this point, the following model is presented to stimulate discourse on the brain basis of parenting behaviors. First, key parenting sensory signals, including cry, visuals as well as touch and smell from baby (A) activate in parallel a set of cortico-limbic circuits (B) to (i) analyze the sensory input and update saliences toward motivation and reward and coordination of other modules for (ii) reflexive caring, (iii) emotion regulation, and (iv) complex cognitions, including mentalization, empathy and theory of mind. The output (C) of these modules forms the basis of parental sensitivity and influences child development. This inclusive and general model may be dissected in future studies involving different stimuli and specific measures of behavior and cognition.

amygdala and the paraventricular nucleus of the hypothala- human mothers is notably less robust, but also supports a mus (Champagne et al., 2001). Following birth, the female role for OT during pregnancy. Mothers with a rising pattern of offsprings of “low” licking and grooming mothers, who are plasma OT reported higher maternal fetal attachment in the cross- fostered and reared by “high” licking and grooming postpartum (Levine et al., 2007), and higher plasma OT levels mothers, express a higher density of brain OT receptors than during pregnancy and the first postpartum month were non-cross-fostered females, and similar levels to high licking correlated with higher levels of maternal postpartum beha- and grooming mothers – suggesting that there is intergenera- viors toward infants, such as gaze, vocalizations and positive tional transmission of the behavior sensitive to the rearing affect in new mothers (Feldman et al., 2007). Maternal syn- environment and likely dependent on epigenetic processes chrony (“episodes when mother and infant coordinate their (Champagne, 2008). positive social engagement”) is also reported to be positively Evidence for a similar “transition to maternity” affected by correlated with maternal plasma OT level, while maternal hormonal exposure during pregnancy and childbirth in intrusiveness (“inappropriate response from mother”) is not brain research 1580 (2014) 78–101 91

(Atzil et al., 2011). OT plasma levels have also been examined oxytocin are currently being investigated in other neuropsychia- in relation to maternal own-attachment experience tric disorders associated with poor social cognition, including (Strathearn et al., 2009): higher levels of plasma OT were ASD, OCD and schizophrenia (e.g. Bartz et al., 2011; Macdonald reported following mother–infant physical interaction among and Feifel, 2013). However, in order to understand the potential mothers reporting secure attachment patterns with their own effects of exogenous oxytocin administration in humans, several mothers compared to those with insecure attachment pat- important lessons from decades of animal research bear con- terns. Even among non-parents, plasma OT levels have been sideration. First, oxytocin effects are moderated by robust, hard- positively correlated with self-reported recall of parental care wired neurobiological mechanisms; thus, centrally administered (maternal and paternal care) (Feldman et al., 2012; Gordon oxytocin increases aggression to same-sex intruders after mating and Feldman, 2008). in monogamous Prairie voles, but not in non-monogamous The literature also points to a role for OT in the regulation Montane voles (Winslow et al., 1993). Similarly, preexisting of stress responses. Indeed, a well-established view has “primers” are necessary to induce oxytocin-facilitated social emerged that, in humans, OT is anxiolytic and reduces fear bonding in monogamous Prairie voles, and, depending on the and stress (Ayers et al., 2011; Ishak et al., 2011). Studies imply study designs, these primers can be gonadal hormones (e.g. that this role is influenced by an individual's previous estrogen), mating behaviors, or prolonged preexposure to a experiences and difficulties in interpersonal relationships former stranger (for an early review, (Insel, 1997)). In other words, (Tabak et al., 2011): difficulties in relationships with primary exposure to oxytocin alone is not sufficient to create de-novo partner (Taylor et al., 2006), or own infant (i.e. interactive social bonding. It seems likely that oxytocin facilitates the stress) (Feldman et al., 2011), or romantic partner (Marazziti development of social bonds only when acting in the appropriate et al., 2006). These studies have all reported higher levels of environment. cross-sectional plasma or urinary OT in relation to stress in These findings have several critical implications for social relationships. Some have concluded that OT appears to humans. Oxytocin administration is likely to affect individuals be an indicator of social affiliation, but it might also be a according to their existing social cognition and attachment “signal” for the need to affiliate with others (Taylor et al., styles as well as their past experiences, much like with the 2010). monogamous Prairie vole and non-monogamous Montane Overall, a large and robust animal literature exists on OT's vole. It is unlikely that acute or chronic exogenous oxytocin role in maternal care behavior and an accumulating, albeit administration alone can reverse genetically and/or environ- weaker human literature associates plasma OT with a wide mentally determined affiliation/attachment styles. Oxytocin range of social and emotional stimuli; scenarios, ranging administration is likely to affect individuals according to their from romantic love, or marital distress to psychopathology. existing social cognition and attachment styles as well as their Interactions between central OT and dopamine systems have past experiences, much like with the monogamous Prairie vole also been associated with individual differences in maternal and non-monogamous Montane vole. So, although perfor- behavior in rodents. For example, stable, individual differ- mance on tasks highlighting empathy (such as trust games) ences in rat maternal licking/grooming of pups were abol- can be enhanced by acute OT administration and empathic ished by OT receptor blockade mediated by the direct effect of interactions can similarly increase plasma OT levels and OT on dopamine release within the mesocorticolimbic dopa- subsequent generosity in humans (Barraza and Zak, 2009), mine system (Shahrokh et al., 2010). While animal models of cultivating empathy in an “uncaring” individual is likely to be maternal behavior are compelling and have identified key a more complex process; not least because its maintenance regions in a putative mothering circuit, human behaviors are will surely rely on the accumulation of repeated positive undertaken in a far more complex environment. As outlined experience. In support of this, oxytocin administration exerts above, processes such as cognitive flexibility, attention con- no significant behavioral change in strangers who share no trol, working memory, and the mother's ability to understand preexisting bonding relationship (Barraza et al., 2011; De Dreu, the intentions and emotions of her child (maternal mind- 2012b; Kosfeld et al., 2005; Zak et al., 2007) and a recent report mindedness or empathy) are fundamental components of finds that intranasal OT does not enhance approach/avoid- human mothering and highly dependent on PFC. Addition- ance to social stimuli or exert a stronger effect on social vs. ally, several studies mentioned already above (Atzil et al., non-social stimuli in the context of processing emotional 2011, 2012; Riem et al., 2011; Strathearn et al., 2009) have information but increases the salience of certain social stimuli attempted to assess neural circuitry related to parenting in and moderates salience of disgust stimuli (Theodoridou et al., subcortical reward/limbic regions as a function of plasma 2013). Authors postulated that heightened responses to dis- oxytocin levels. gust may be particularly relevant for new mothers wishing to The clinical implications of these studies are yet to be fully protect vulnerable young from contagion. explored. However, preliminary observations that expectant The positive effects of exogenous OT on social cognition mothers at risk for postpartum depression have lower plasma may only be observed when the tasks demand a binding OT during pregnancy (Skrundz et al., 2011)maysuggestvaluein relationship in which reward contingencies are congruent interventions which enhance availability of central OT during between two players, where they form a partnership by pregnancy and perinatally in order to reduce risk of postpartum becoming stake holders of common interest (Barraza et al., depression. Perhaps certain administration regimes may be 2011; Kosfeld et al., 2005; Zak et al., 2007).Oxytocin adminis- explored to overcome likely problems with the overlapping role tration also appears to increase polarized social behaviors in of OT in uterine contraction during pregnancy and the risk humans, e.g., in-group favoritism (De Dreu et al., 2011), by therefore of inducing premature labor. Therapeutic uses of increasing within group cooperation and between group 92 brain research 1580 (2014) 78–101

competition (De Dreu, 2012a) and its administration may 3.1. Controversies and limits to the OT literature in increase the emotional reactivity to perceived positive and parenting negative cues in a social context (Olff et al., 2013). When social cues in the environment are interpreted as secure or The nature of the interaction between peripheral and central OT positive (“me among us”), oxytocin may promote prosocial systems moment-to-moment remains elusive. Recent evidence behaviors. In situations of higher threat or higher social in 41 non-psychiatric subjects concluded that there is no stress, when the social cues are interpreted as insecure or correlation between baseline CSF and plasma OT levels mea- negative (“me among them”), oxytocin may promote “anti- sured using sensitive radioimmunoassay (Kagerbauer et al., social” behaviors. This suggests that exogenous OT may serve 2013). These authors rigorously appraise past literature using to sharpen in/out group distinction (Colonnello et al., 2013). basal measurements of peripheral OT to reflect central effects of Interestingly, it was found that OT levels in mothers, the hormone in humans, particularly in studies proposing links assayed from urine, were higher after interaction with chil- between plasma/salivary OT and human social behaviors. Such a dren, no matter whether these were their own or unknown critique of the evidence must caution future studies to acknowl- children (Bick and Dozier, 2010); also see Elmadih et al., 2014. edge that OT occupies two physiologically distinct compart- In fact, OT levels appeared to be higher if the child was ments (periphery and CSF) resulting in quite independent unknown compared to when the child was their own – release patterns. This suggests that measuring plasma OT as a perhaps serving an anxiolytic or anti-stress function in this biomarker of a socio-emotional behavior assumed to be regu- particular experiment (Heinrichs et al., 2003). It may also fit lated by brain perhaps remains relevant only if a dynamic with a role for OT in maternal protection and aggression challenge paradigm is used with comparisons between distinct, towards intruders in several species including humans well-characterized groups. Kagerbauer et al. (2013) acknowledge (Campbell, 2008; Neumann, 2008). Finally, in articles from the relationship between gender, gonadal hormones and plasma this special issue, OT was measured and administered in OT; but they also point to a high quality study in pregnant respective experiments on mother behavior and non-mother women showing no correlation between plasma and CSF OT responses to infant pictures. New mothers with low sensi- levels (Altemus et al., 2004). tivity showed higher plasma OT before and after playing with Understanding how intranasal oxytocin affects human their infants compared to mothers with higher than average social behaviors clearly requires more thoughtful experimen- maternal sensitivity (Elmadih et al., 2014). These data suggest tation. These considerations do not foreclose the potential for that OT is released as part of a stress response related to poor exogenously/intranasally administered OT to reach brain and ability to cope with infant demands and may be moderating affect behavior relevant to parenting. Indeed, intranasal drug maternal stress in an attempt to promote care and bonding. administration is increasingly favored as an effective means This is complemented by the report that amygdala response of delivering molecules rapidly to brain (Grassin-Delyle et al., to infant pictures among nulliparous women are enhanced by 2012). But it is likely that effects are not only sex-specific, but intranasal OT yet with more incorrect facial expression task specific (some suggest that OT does not change the interpretation (Voorthuis et al., 2014) – again suggesting a salience of social–emotional cues) and task-specific between role for OT in enhanced coping that may even interfere with sexes (Ditzen et al., 2012). Beyond pregnancy and infancy, nuanced responses to infant stimuli – although findings may we have little information on how the OT system func- be different among parents or with different OT administra- tions in socio-emotional relationships during periods of tion protocols. rapid brain reorganization (e.g. puberty) vs. those of relative The complex literature on the effects in humans of developmental stability and we know little in humans about externally administered hormone argue for further refine- brain OT-receptor distributions and densities and how they ments in our understanding of how, and in what direction OT may be influenced by environmental effects and epigenetic (measured in the plasma or elsewhere) is, or is not, causally regulation overtime and circumstance. To what extent cen- associated with behavioral or emotional manifestations in tral OT receptors are sensitive to epigenetic modification in people. When someone inhales OT, evidence points to its humans and whether, like in the rodents from Champagne's effects being determined by past experiences and present (2008) work, they contribute to patterns of parental care that circumstances. For example, we observe that intranasal can be passed on (in a quasi-Lamarckian fashion) from oxytocin may increase startle responses to stressful stimuli generation to generation, are closely related questions. Other in humans (Grillon et al., 2013; Striepens et al., 2012) and that hormones and their receptor systems, including cortisol, less anxiously attached, healthy men remember their mother estrogen, progesterone, endogenous opioids, are also impor- as more caring and close after OT (vs. placebo), while more tant in the regulation of parenting and interact with OT anxiously attached men remember their mother as less (Gordon et al., 2011; Lahey et al., 2012; Nowak et al., 2011; caring and close after OT (vs. placebo) (Bartz et al., 2010). Swain et al., 2011) Finally, healthy men with low emotion regulation abilities In summary, the complex links between OT and parental show higher cortisol stress responses (as expected), but also brain physiology in humans requires further, more detailed benefit more from intranasal OT; whereas those with high examination, taking into account key individual differences emotion regulating abilities do not (Quirin et al., 2011). This and maternal experiences past and present into account, in chimes well with early studies (Light et al., 2000) that addition to other hormones, in order to appropriately inter- reported increased plasma OT in mothers following a social pret the inferences that can be drawn from studies using stress task only when the mother held her baby before the basal peripheral measures of the hormone alongside beha- task, but not if she did not. vioral correlates. brain research 1580 (2014) 78–101 93

mother–infant interactions (Dickson et al., 1997; Feldman, 4. Advances on the neurobiology of father 2003; Lamb, 1977; Yogman, 1981). Thus, unique contributions brain from father–child interactions (Boyce et al., 2006; Grossmann et al., 2002) to evolutionarily favorable sex-specific emotional Studies of parent–infant interactions have historically tar- expressions of the developing child may significantly consti- geted the mother–infant relationship. However, more recent tute the mechanism through which sex differences in the research has demonstrated that fathering also plays a sig- expression of emotion (Vigil, 2009) cross generations. nificant role for the child's cognitive, emotional and social The differential nature of father– vs. mother–child rela- development (Lamb, 2004; Ramchandani and Psychogiou, tionship suggests that the underlying neurobiology of father- 2009). For example, literature examining parental behaviors ing behaviors may be at least in part distinct. For example, in suggests gender differences in expressed emotion during the model presented here (Fig. 1), paternal salience appraisal interactions with children (Volling et al., 2002). Specifically, of infant stimuli may activate an emotional regulatory reac- maternal sensitivity is typically expressed by emotional tion in fathers that is stronger than that evoked for mothers. warmth and support whereas paternal sensitivity frequently This kind of a response would underlie paternal capacity and manifests as the provision of stimulating interactions motivation to effectively assess external contextual factors to (Grossmann et al., 2008). This gender difference is supported determine whether situations are sufficiently safe to encou- by a study of a time series analysis of 100 first-time mothers rage the child's engagement and interactions with the and fathers interacting with their 5-month-old firstborn child broader social environment. In contrast, hormonal changes (Feldman, 2003). Mother–child play was characterized by face- corresponding to the perinatal period may result in a stronger to-face exchange and included patterns of mutual gazing, reflexive caring response for mothers. Finally, empathic/ covocalization, and affectionate touch. In contrast, during mentalizing cognitive responses to infant cues may be play with fathers, a pattern of interactive arousal was similar for mothers and fathers, as it supports all types of identified that contained several quick peaks of high positive parenting responses and behaviors – at least after an initial emotionality, including joint laughter and open exuberance. period of adaptation of some months. This is not well studied Behavioral data with fathers have demonstrated impor- but at least in keeping with the observation mother vs. father tant differences in typical father vs. mother–child interac- statistical equivalence of correct identification of own vs. tions beginning in infancy (Crawley and Sherrod, 1984; other baby-cry (Swain et al., 2005). Feldman, 2003). Fathers, for example, tend to exhibit Preliminary work on the brain physiology of parents is increased physical interactions with their young children, consistent with significantly overlapping neurobiological and often characterized as “rough and tumble” play (Carson et al., behavioral parenting processes for mothers and fathers with 1993). Furthermore, within father–child relationships, it is the some interesting differences. For example, using the fMRI quality of active, parent–child play interactions, and not methodology with a sample of coparent couples of 4–6- sensitivity per se, that is related to positive social emotional month old infants, Atzil et al. (2012) found greater activation outcomes in children such as interpersonal confidence and in limbic areas in response to own-infant (vs. other infant) the development of positive peer relationships (MacDonald, video clips for mothers (and not fathers) as well as correla- 1987). These data suggest that there may be some shared tions of these areas (e.g., NAcc, amygdala, and ventral ACC) brain mechanisms between mothers and fathers, and some with increases in plasma oxytocin (OT) levels. In contrast, differences. Recently, an interesting study with small groups decreases in plasma OT for fathers were correlated in own- vs. showed that males respond differently to non-own hunger other-infant comparisons, primarily with cognitive areas, cries than mothers (De Pisapia et al., 2013). Studying 9 men including areas that are responsible for regulating and orga- and 9 women (half in each group non-parents), more activity nizing behavioral responses to emotionally salient stimuli was seen in the dorsal medial prefrontal and posterior (e.g., dorsolateral PFC, dorsal ACC, IPC, and precentral gyrus). cingulate areas of male brains regardless of parental status. Further, in fathers (and not mothers), increases in plasma Such sex differences were interpreted as relating to mind- arginine vasopressin (AVP) were correlated with activations wandering in men relative to women. Sex differences of the inferior frontal gyrus (IFG) and insula suggesting an between mothers and fathers may relate to the establish- increased social-cognitive response to infant cues for fathers. ment and maintenance of specific aspects of parent–infant Using similar methodology with a sample of young infants relationship and warrant a further study. A theoretical basis (8–19 weeks), the brain responses of fathers to own vs. other for sex differences may be found in differential roles of infant videos (Kuo et al., 2012) were obtained. Whole brain mothers and fathers in the development of exploratory analysis demonstrated increased activity in emotion regula- systems in the infant and young child according to activation tion (iii) circuits, including bilateral inferior frontal gyrus, and relationship theory (Paquette, 2004). This theory includes two the empathic/mentalizing (iv) module including supramargi- dimensions of fathering that underlie the father–child rela- nal (parietal) gyrus and bilateral middle temporal gyrus tionship: (1) stimulation, wherein fathers encourage the child's among human fathers (N 10) in response to own (vs. other) ¼ interaction with the outside world; and (2) discipline designed baby stimuli. In addition, on the contrast of baby (both own to provide children with limits that will maintain their safety and other babies) vs. doll video contrasts, fathers exhibited (Paquette, 2004). increased activity in all modules (Fig. 1): Sensory/Salience (i), Such specialization of father–infant interactions is consis- Reflexive Caring (ii), Emotion regulation (iii), and Empathic/ tent with play studies showing more object-oriented or Mentalizing (iv), including bilateral caudate, orbitofrontal physical play associated with smiles compared with cortex, superior frontal gyrus, and superior . 94 brain research 1580 (2014) 78–101

Activation of these systems plausibly supports a father's ability that fit with a picture of paternal brain-behavior adaptations to attend to the infant and the surrounding environment and coordinated with relevant hormones after having a child in react as necessary to promote infant–environment interactions support of adaptive parenting. Future work will no doubt or to protect the infant from environmental dangers. Interest- explore similarities and differences to the maternal brain. ingly, greater responses to own (vs. other) baby stimuli in the PFC were associated with less paternal sensitivity. The findings may reflect the greater role of the OFC in interpretation of the 5. Conclusions and future directions unfamiliar infant cry among fathers. Using infant photographs instead of video clips to stimu- By mapping parental brain responses to infant stimuli, non- late father brains, Wittfoth-Schardt et al. (2012) found activa- invasive and highly sensitive functional magnetic resonance tions for own child in the left GP, the left hippocampus, the imaging (fMRI) has added greatly to our knowledge of the right mOFC (ii & iv, Fig. 1), as well as in the bilateral inferior brain basis of parental sensitivity. Furthermore, it has been frontal gyrus/anterior insula (iv) and activations for own child confirmed that the observed maternal behavior reflects a vs. other (unknown) child in the right GP, the left VTA (ii), the physiological composite of multiple behaviors, with discrete left mOFC, and the left inferior frontal gyrus(IFG)/anterior maternal brain activation likely to occur in relation to each insula (iv). Additionally, OT administration reduced activa- aspect of this behavior. Such conceptualization presents the tion and functional connectivity of the left GP with reward- possibility of identifying distinct neuro-hormonal pathways and attachment-related regions responsive to pictures of to poor maternal sensitivity, and of using changes in brain both own and unfamiliar child. Furthermore, in recent ana- activation in response to infant stimuli as potential biomar- lyses of father brain activity, using an own vs. other child vs. kers for the development and evaluation of new diagnostic adult picture task 1–2 years postpartum (Mascaro et al., and treatment strategies in at-risk parents. 2013b), there was a main effect for child vs. adult pictures Indeed, there are now many examples in which brain in the fusiform gyrus, dorsal medial prefrontal cortex, thala- imaging can identify changes in neuronal activation that mocingulate and mesolimbic areas – fitting with arousal, elude behavioral measures. For example amygdala activation reflexive caring and emotion regulation parental brain cir- and connectivity is significantly increased during subcon- cuits (Fig. 1). VTA responses were associated with reduced scious processing of fearful faces (Pantazatos et al., 2012). testosterone levels, smaller testes volume as well as higher fMRI might then provide insight into the neural basis of parental sensitivity. The findings support that fathers may subconscious or “automatic” parental responses (Papousek gradually develop a strong attachment with their infants over and Papousek, 2002). Similarly, behavioral measures of emo- the course of the first and second years, and the attachment tional function can be relatively insensitive, whereas using building processes may be supported by neurobiological affective cognition challenges, fMRI has allowed the identifi- adaptations such as increased neural sensitivity to own cation of important biomarkers for psychopathology and for infants and decreased testosterone levels. In response treatment response in depression such as the amygdala and to 3–5 month-old infant cries, similar to the findings in ACC (Groenewold et al., 2012; Harmer et al., 2009, 2011). In mothers were reported for fathers 1–2 years postpartum this way, imaging data complements and enhances beha- (Mascaro et al., 2013a), including responses in bilateral IFG vioral information. For example, observational measures of and GP. Furthermore, variations in androgen receptor gene behavior can address questions about whether a parent at were associated with activations in IFG and OFC – areas also risk of poor maternal care (or poor maternal sensitivity) has involved in empathy and emotion regulation (iii & iv, Fig. 1). biases in attention systems, but only imaging can address Anterior insula activity had a non-linear relationship with questions about how such a parent attributes salience or paternal caregiving, such that fathers with intermediate valence to emotional images of her own as compared to an activation were most involved. These results suggest that unknown infant, and how they might differentially engage restrictive attitudes may be associated with decreased empa- brain reward systems. A combination of brain imaging plus thy and emotion regulation in response to a child in distress, behavioral paradigms seems more likely to result in and that moderate anterior insula activity reflects an optimal enhanced mechanistic specificity. Thus, if we find impaired level of arousal that supports engaged fathering. performance in a parent behaviorally when he or she is under Finally, the father brain structure is also being examined stress, then the imaging may be able to indicate if this is based on animal models (Kinsley and Lambert, 2008), and one because of over-activity in limbic stress systems (e.g. neuroimaging report on gray matter volume increases in a enhanced amygdale and insula activation to stressful cues) number of brain regions of human mothers – including the and/or because prefrontal control systems are underactive, striatum, thalamocingulate, and PFC from the first month (T1) to making executive performance more readily disrupted. Hav- the fourth postpartum month (T2) (Kim et al., 2010a). For new ing more information about which neural mechanisms drive fathers, gray matter changes may be associated with depression which behavioral outcome has important implications for symptoms and paternal behaviors (Kim et al., submitted for designing more effective interventions – e.g. depending on publication). Ultimately, models that combine changes in struc- the parent, interventions may focus on reducing stress ture and function may be required for comprehensive biological responses, or improving executive performance, or both. models of paternal behavior that may help identify biomarkers The interpretation of fMRI data on parents' brains for risk, resilience and intervention. responding to infant stimuli carries important caveats. These In summary, brain imaging studies of the father brain over include consideration of timing, averaging of signals and the last few years have yielded functional and structural results limitation of the paradigms themselves, which represent a brain research 1580 (2014) 78–101 95

set of circumstances quite different from those actually determine whether low sensitivity parents show reduced experienced by a human parent in real-time with their own reward system responses to their infants – an extremely infant (Hari and Kujala, 2009; Schilbach et al., 2013). Lying in a challenging task based on observed behaviors or self-report. magnet observing audiovisual stimuli is far from actual In the same way, within-subject, placebo-controlled designs parenting and the field is working toward more ethologically combined with the superior sensitivity of imaging are likely accurate paradigms for use in brain imaging experiments of to be of particular value in answering whether acute oxytocin parental responsiveness. can modulate affective cognition and other neural responses In spite of these caveats and need for replication, fMRI to infants in low-sensitivity mothers. studies of parental brain have revealed a set of cortico-limbic Future fMRI investigation of parental sensitivity and its brain circuits that are activated in response to infant stimuli, modulation should include examination of the relationship and some of which track objectively measures parental between deficits in key affective cognition pathways. Increas- sensitivity (Kim et al., 2011; Musser et al., 2012). Since ing amounts of money are likely to be spent on attempts to parental behavior necessarily changes as infants grow and improve parenting or reduce the adverse effects of poor mature, so too do we have evidence that brain imaging parenting. In our opinion, the success of such approaches will activations change over time among healthy new mothers continue to be limited unless we understand in detail the (Kim et al., 2010a) Plausible next research steps could include neurobiology which underpins poor parental sensitivity. This studies of the biological mechanisms required for specific means being able to distinguish differing pathways to, and aspects of parental brain adaptation. Such approaches to components of, poor sensitivity. Brain imaging represents a brain plasticity may be fruitful to understand the pathophy- highly efficient methodology to explore this in relatively small siology of parent-infant relational problems as well as poten- numbers compared to numbers needed even for pilot rando- tial opportunities for risk/resilience identification and brain- mized control trials using exclusively behavioral outcomes. based intervention research. Intranasal oxytocin remains an Significant differences in neural activation with intervention interesting candidate to help augment parenting in future vs. placebo, (even if behavioral change is not recorded), would therapies. still support the value of comparing a combination of interven- Over the past decade, a number of clinical studies demon- tion plus behavioral vs. intervention alone in a future potential strate the positive effect of intranasal OT (exogenous OT) future randomized controlled trial. on emotion recognition/social anxiety (e.g. (Guastella and MacLeod, 2012) and affiliative behavior between individuals (Riem et al., 2011), including fathers (Weisman et al., 2012). Acknowledgments These suggest that OT administration may be indicated for vulnerable parents with poor maternal sensitivity, although The authors of this paper are supported by grants from the caution is still required (Tabak et al., 2011). If fMRI can National Alliance for Research on Schizophrenia and Depres- discriminate different patterns of brain activation between sion/Brain and Behavior Foundation (JES), the Klingenstein parents at opposite ends of a spectrum of high and low Third Generation Foundation, NIMHD/NICHD RC2MD004767- fi maternal sensitivity, it prepares the way for future ef cient 01, the Michigan Institute for Clinical Health Research, the hypothesis testing of the effects of novel interventions in National Center for Advancing Translational Sciences, small numbers of normal volunteers. In other words, a UL1TR000433 (JES, SSH, and CJD), Robert Wood Johnson fi “ ” distinct neural pro le of higher sensitivity parents or Health and Society Scholar Award (JES and SSH) and NIMH parenting response means functional imaging may identify 5T32MH018264-28 (JS). useful “biomarkers” for future interventions; for example, monitoring the effects of intranasal OT or non-drug inter- references ventions aimed at enhancing parental responsiveness on aspects of neural circuitry. We would argue that fMRI utility derives from the propo- Abel, K.M., Webb, R.T., Salmon, M.P., Wan, M.W., Appleby, L., 2005. fi sal that speci c fMRI activation patterns could act as objec- Prevalence and predictors of parenting outcomes in a cohort tive biomarkers for treatment development and in risk of mothers with schizophrenia admitted for joint mother and identification. We would suggest that future work using fMRI baby psychiatric care in England. J. Clin. Psychiatry 66, 781–789 in new parents should assess affective cognition and how it (quiz 808–789). relates to parental care behaviors before and after interven- Ainsworth, M.S., Blehar, M.C., Waters, E., Wall, S., 1978. In: Patterns of Attachment: A Psychological Study of the Strange tions in order to determine underlying neural pathways for Situation. Erlbaum, Oxford, England. “ ” what is, and what is not working with respect to the Altemus, M., Fong, J., Yang, R., Damast, S., Luine, V., Ferguson, D., parental care. Such a model is already being developed for 2004. Changes in cerebrospinal fluid neurochemistry during the mood disorders, in which fMRI studies have consistently pregnancy. Biol. Psychiatry 56, 386–392. demonstrated that limbic hyperactivity in response to emo- Anderson, A.M., 1996. Factors influencing the father–infant tional faces offers promise as a biomarker for response to relationship. J. Fam. Nurs. 2, 306–324. anti- (Harmer et al., 2011; Victor et al., 2012). Anderson, I.M., Shippen, C., Juhasz, G., Chase, D., Thomas, E., Downey, D., Toth, Z.G., Lloyd-Williams, K., Elliott, R., Similarly, abnormalities in response to rewarding stimuli Deakin, J.F., 2011. State-dependent alteration in face emotion fi are dif cult to access in human behavioral studies, but recognition in depression. Br. J. Psychiatry 198, 302–308. numerous fMRI studies have mapped the different neural Arnone, D., McKie, S., Elliott, R., Thomas, E.J., Downey, D., responses to different rewards. Brain imaging may be able to Juhasz, G., Williams, S.R., Deakin, J.F.W., Anderson, I.M., 2012. 96 brain research 1580 (2014) 78–101

Increased amygdala responses to sad but not fearful faces in biological and non-biological children. Dev. Psychobiol. 52, major depression: relation to mood state and pharmacological 100–107. treatment. Am. J. Psychiatry 169, 841–850. Born, J.M., Lemmens, S.G., Martens, M.J., Formisano, E., Goebel, R., Atkinson, L., Leung, E., Goldberg, S., Benoit, D., Poulton, L., Westerterp-Plantenga, M.S., 2011. Differences between liking Myhal, N., Blokland, K., Kerr, S., 2009. Attachment and and wanting signals in the human brain and relations with selective attention: disorganization and emotional Stroop cognitive dietary restraint and body mass index. Am. J. Clin. reaction time. Dev. Psychopathol. 21, 99–126. Nutr. 94, 392–403. Atzil, S., Hendler, T., Feldman, R., 2011. Specifying the Bos, P.A., Hermans, E.J., Montoya, E.R., Ramsey, N.F., van Honk, J., neurobiological basis of human attachment: brain, hormones, 2010. Testosterone administration modulates neural and behavior in synchronous and intrusive mothers. responses to crying infants in young females. Neuropsychopharmacology 36, 2603–2615. Psychoneuroendocrinology 35, 114–121. Atzil, S., Hendler, T., Zagoory-Sharon, O., Winetraub, Y., Feldman, Bowlby, J., 1969a. In: Attachment 2nd ed. Basic Books, New York. R., 2012. Synchrony and specificity in the maternal and Bowlby, J., 1969b. Attachment and loss. Attachment, vol. 1. the paternal brain: relations to oxytocin and vasopressin. Hogarth Press, London. J. Am. Acad. Child Adolesc. Psychiatry 51, 798–811. Bowlby, J., 1973. Attachment and loss. Separation: Anxiety and Ayers, L.W., Missig, G., Schulkin, J., Rosen, J.B., 2011. Oxytocin Anger, vol. 2. Basic Books, London. reduces background anxiety in a fear-potentiated startle Bowlby, J., 1977. The making and breaking of affectional bonds. I. paradigm: peripheral vs central administration. Aetiology and psychopathology in the light of attachment Neuropsychopharmacology 36, 2488–2497. theory. An expanded version of the Fiftieth Maudsley Lecture, Bakermans-Kranenburg, M.J., van, I.M.H., Juffer, F., 2003. Less is delivered before the Royal College of Psychiatrists, 19 more: meta-analyses of sensitivity and attachment November 1976. Br. J. Psychiatry 130, 201–210. interventions in early childhood. Psychol. Bull. 129, Boyce, W.T., Essex, M.J., Alkon, A., Goldsmith, H.H., Kraemer, H.C., 195–215. Kupfer, D.J., 2006. Early father involvement moderates Barraza, J.A., McCullough, M.E., Ahmadi, S., Zak, P.J., 2011. biobehavioral susceptibility to mental health problems in Oxytocin infusion increases charitable donations regardless of middle childhood. J. Am. Acad. Child Adolesc. Psychiatry 45, monetary resources. Horm. Behav. 60, 148–151. 1510–1520. Barraza, J.A., Zak, P.J., 2009. Empathy toward strangers triggers Bridges, R.S., 2008. In: Neurobiology of the Parental Brain. oxytocin release and subsequent generosity. Ann. N. Y. Acad. Academic Press, Burlington, MA, USA. Sci. 1167, 182–189. Britton, J.C., Phan, K.L., Taylor, S.F., Welsh, R.C., Berridge, K.C., Barrett, J., Fleming, A.S., 2011. Annual research review: all Liberzon, I., 2006. Neural correlates of social and nonsocial mothers are not created equal: neural and psychobiological emotions: an fMRI study. Neuroimage 31, 397–409. perspectives on mothering and the importance of individual Campbell, A., 2008. Attachment, aggression and affiliation: the differences. J. Child Psychol. Psychiatry 52, 368–397. role of oxytocin in female social behavior. Biol. Psychol. 77, Barrett, J., Wonch, K.E., Gonzalez, A., Ali, N., Steiner, M., Hall, G.B., 1–10. Fleming, A.S., 2012. Maternal affect and quality of parenting Caria, A., Falco, S., Venuti, P., Lee, S., Esposito, G., Rigo, P., experiences are related to amygdala response to infant faces. Birbaumer, N., Bornstein, M.H., 2012. Species-specific Soc. Neurosci. 7, 252–268. response to human infant faces in the premotor cortex. Bartels, A., Zeki, S., 2004. The neural correlates of maternal and Neuroimage 60, 884–893. romantic love. Neuroimage 21, 1155–1166. Carson, J., Burks, V., Parke, R.D., 1993. Parent–child physical play: Bartz, J.A., Zaki, J., Bolger, N., Ochsner, K.N., 2011. Social effects of determinants and consequences. In: MacDonald, K. (Ed.), oxytocin in humans: context and person matter. Trends Cogn. Parent–Child Play: Descriptions and Implications. State Sci. 15, 301–309. University of New York Press, Albany, NY, USA, pp. 197–220. Bartz, J.A., Zaki, J., Ochsner, K.N., Bolger, N., Kolevzon, A., Ludwig, Champagne, F., Diorio, J., Sharma, S., Meaney, M.J., 2001. Naturally N., Lydon, J.E., 2010. Effects of oxytocin on recollections of occurring variations in maternal behavior in the rat are maternal care and closeness. Proc. Natl. Acad. Sci. USA 107, associated with differences in estrogen-inducible central 21371–21375. oxytocin receptors. Proc. Natl. Acad. Sci. USA 98, 12736–12741. Baxter Jr., L.R., 2003. Basal ganglia systems in ritualistic social Champagne, F.A., 2008. Epigenetic mechanisms and the displays: reptiles and humans; function and illness. Physiol. transgenerational effects of maternal care. Front. Behav. 79, 451–460. Neuroendocrinol. 29, 386–397. Behrens, K.Y., Hart, S.L., Parker, A.C., 2012. Maternal sensitivity: Champagne, F.A., 2010. Epigenetic influence of social experiences evidence of stability across time, contexts, and measurement across the lifespan. Dev. Psychobiol. 52, 299–311. instruments. Infant Child Dev. 21, 348–355. Champagne, F.A., Francis, D.D., Mar, A., Meaney, M.J., 2003. Belsky, J., Jaffee, S.R., Sligo, J., Woodward, L., Silva, P.A., 2005. Variations in maternal care in the rat as a mediating influence Intergenerational transmission of warm-sensitive-stimulating for the effects of environment on development. Physiol. parenting: a prospective study of mothers and fathers of 3- Behav. 79, 359–371. year-olds. Child Dev. 76, 384–396. Clutton-Brock, T.H., 1991. In: The Evolution of Parental Care. Benedek, 1954. Parenthood as a developmental phase: a contribution Princeton University Press; Princeton, NY, USA. to the libido theory. J. Am. Psychoanal. Assoc. 7, 379–417. Colonnello, V., Chen, F.S., Panksepp, J., Heinrichs, M., 2013. Bernhardt, B.C., Singer, T., 2012. The neural basis of empathy. Oxytocin sharpens self-other perceptual boundary. Annu. Rev. Neurosci. 35, 1–23. Psychoneuroendocrinology. Bernier, A., Carlson, S.M., Whipple, N., 2010. From external Crawley, S.B., Sherrod, K.B., 1984. Parent–infant play during the regulation to self-regulation: early parenting precursors of first year of life. Infant Behav. Dev. 7, 65–75. young children’s executive functioning. Child Dev. 81, 326–339. Curley, J.P., 2011. The mu-opioid receptor and the evolution of Berridge, K.C., Kringelbach, M.L., 2008. Affective neuroscience of mother–infant attachment: theoretical comment on Higham pleasure: reward in humans and animals. et al. (2011). Behav. Neurosci. 125, 273–278. Psychopharmacology 199, 457–480. De Dreu, C.K., 2012a. Oxytocin modulates cooperation within and Bick, J., Dozier, M., 2010. Mothers’ and children’s concentrations competition between groups: an integrative review and of oxytocin following close, physical interactions with research agenda. Horm. Behav. 61, 419–428. brain research 1580 (2014) 78–101 97

De Dreu, C.K., 2012b. Oxytocin modulates the link between adult Frith, C.D., Frith, U., 2006. The neural basis of mentalizing. attachment and cooperation through reduced betrayal Neuron 50, 531–534. aversion. Psychoneuroendocrinology 37, 871–880. Frith, U., Frith, C.D., 2003. Development and neurophysiology of De Dreu, C.K., Greer, L.L., Van Kleef, G.A., Shalvi, S., Handgraaf, M.J., mentalizing. Philos. Trans. R. Soc. Lond. B: Biol. Sci. 358, 2011. Oxytocin promotes human ethnocentrism. Proc. Natl. 459–473. Acad. Sci. USA 108, 1262–1266. Galbally, M., Lewis, A.J., Ijzendoorn, M., Permezel, M., 2011. The De Pisapia, N., Bornstein, M.H., Rigo, P., Esposito, G., De Falco, S., role of oxytocin in mother-infant relations: a systematic Venuti, P., 2013. Sex differences in directional brain responses review of human studies. Harv. Rev. Psychiatry 19, 1–14. to infant hunger cries. Neuroreport 24, 142–146. Gallese, V., Keysers, C., Rizzolatti, G., 2004. A unifying view of the De Wolff, M., van Ijzendoorn, M.H., 1997. Sensitivity and basis of social cognition. Trends Cogn. Sci. 8, 396–403. attachment: a meta-analysis on parental antecedents of Glocker, M.L., Langleben, D.D., Ruparel, K., Loughead, J.W., Valdez, infant attachment. Child Dev. 68, 571–591. J.N., Griffin, M.D., Sachser, N., Gur, R.C., 2009. Baby schema Dickson, K.L., Walker, H., Fogel, A., 1997. The relationship modulates the brain reward system in nulliparous women. between smile type and play type during parent–infant play. Proc. Natl. Acad. Sci. USA 106, 9115–9119. Dev. Psychol. 33, 925–933. Gonzalez, A., Jenkins, J.M., Steiner, M., Fleming, A.S., 2012. Ditzen, B., Nater, U.M., Schaer, M., La Marca, R., Bodenmann, G., Maternal early life experiences and parenting: the mediating Ehlert, U., Heinrichs, M., 2012. Sex-specific effects of role of cortisol and executive function. J. Am. Acad. Child intranasal oxytocin on autonomic nervous system and Adolesc. Psychiatry 51, 673–682. emotional responses to couple conflict. Soc. Cogn. Affect. Goodman, S.H., Rouse, M.H., Connell, A.M., Broth, M.R., Hall, C.M., Neurosci.. Heyward, D., 2011. Maternal depression and child Eisenberg, N., Cumberland, A., Spinrad, T.L., Fabes, R.A., psychopathology: a meta-analytic review. Clin. Child Fam. Shepard, S.A., Reiser, M., Murphy, B.C., Losoya, S.H., Psychol. Rev. 14, 1–27. Guthrie, I.K., 2001. The relations of regulation and Gordon, I., Feldman, R., 2008. Synchrony in the triad: a microlevel emotionality to children’s externalizing and internalizing process model of coparenting and parent–child interactions. problem behavior. Child Dev. 72, 1112–1134. Fam. Process 47, 465–479. Eisenberger, N.I., Cole, S.W., 2012. Social neuroscience and health: Gordon, I., Martin, C., Feldman, R., Leckman, J.F., 2011. Oxytocin neurophysiological mechanisms linking social ties with and social motivation. Dev. Cogn. Neurosci. 1, 471–493. physical health. Nat. Neurosci. 15, 669–674. Grassin-Delyle, S., Buenestado, A., Naline, E., Faisy, C., Blouquit- Elliott, R., Zahn, R., Deakin, J.F., Anderson, I.M., 2011. Affective Laye, S., Couderc, L.J., Le Guen, M., Fischler, M., Devillier, P., cognition and its disruption in mood disorders. 2012. Intranasal drug delivery: an efficient and non-invasive Neuropsychopharmacology 36, 153–182. route for systemic administration: focus on opioids. Elmadih, A., Wai Wan, M., Numan, M., Elliott, R., Downey, D., Pharmacol. Ther. 134, 366–379. Abel, K.M., 2014. Does oxytocin modulate variation in Grillon, C., Krimsky, M., Charney, D.R., Vytal, K., Ernst, M., maternal caregiving in healthy new mothers? Brain Res. 2014, Cornwell, B., 2013. Oxytocin increases anxiety to epub ahead of print 1580, 143–150. unpredictable threat. Mol. Psychiatry 18, 958–960. Engfer, A., Gavranidou, M., 1987. Antecedants and consequences Groenewold, N.A., Opmeer, E.M., de Jonge, P., Aleman, A., of maternal sensitivity: a longitudinal study. In: Rauh, H., Costafreda, S.G., 2012. Emotional valence modulates brain Steinhausen, H.-C. (Eds.), Psychobiology and Early functional abnormalities in depression: evidence from a Development, pp. 71–99. meta-analysis of fMRI studies. Neurosci. Biobehav. Rev. 37, Feldman, R., 2003. Infant–mother and infant–father synchrony: 152–163. the coregulation of positive arousal. Inf. Mental Health J. 24, Grossmann, K., Grossmann, K.E., Fremmer-Bombik, E., Kindler, 1–23. H., Scheuerer-Englisch, H., Zimmerman, P., 2002. The Feldman, R., Gordon, I., Zagoory-Sharon, O., 2011. Maternal and uniqueness of the child–father attachment relationship: paternal plasma, salivary, and urinary oxytocin and parent- fathers’ sensitive and challenging play as a pivotal variable in infant synchrony: considering stress and affiliation a 16-year longitudinal study. Soc. Dev. 11, 307–331. components of human bonding. Dev. Sci. 14, 752–761. Grossmann, K., Grossmann, K.E., Kindler, H., Zimmermann, P., Feldman, R., Granat, A., Pariente, C., Kanety, H., Kuint, J., Gilboa- 2008. A wider view of attachment and exploration: the Schechtman, E., 2009. Maternal depression and anxiety across influence of mothers and fathers on the development of the postpartum year and infant social engagement, fear psychological security from infancy to young adulthood. In: regulation, and stress reactivity. J. Am. Acad. Child Adolesc. Cassidy, J., Shaver, P.R. (Eds.), Handbook of Attachment: Psychiatry 48, 919–927. Theory, Research, and Clinical Applications. Guilford Press, Feldman, R., Weller, A., Leckman, J.F., Kuint, J., Eidelman, A.I., New York, NY, pp. 857–879. 1999. The nature of the mother’s tie to her infant: maternal Guastella, A.J., MacLeod, C., 2012. A critical review of the bonding under conditions of proximity, separation, and influence of oxytocin nasal spray on social cognition in potential loss. J. Child Psychol. Psychiatry 40, 929–939. humans: evidence and future directions. Horm. Behav. 61, Feldman, R., Weller, A., Zagoory-Sharon, O., Levine, A., 2007. 410–418. Evidence for a neuroendocrinological foundation of human Hari, R., Kujala, M.V., 2009. Brain basis of human social affiliation: plasma oxytocin levels across pregnancy and the interaction: from concepts to brain imaging. Physiol. Rev. 89, postpartum period predict mother–infant bonding. Psychol. 453–479. Sci. 18, 965–970. Harmer, C.J., Cowen, P.J., Goodwin, G.M., 2011. Efficacy markers in Feldman, R., Zagoory-Sharon, O., Weisman, O., Schneiderman, I., depression. J. Psychopharmacol. 25, 1148–1158. Gordon, I., Maoz, R., Shalev, I., Ebstein, R.P., 2012. Sensitive Harmer, C.J., O’Sullivan, U., Favaron, E., Massey-Chase, R., parenting is associated with plasma oxytocin and Ayres, R., Reinecke, A., Goodwin, G.M., Cowen, P.J., 2009. Effect polymorphisms in the OXTR and CD38 genes. Biol. Psychiatry of acute antidepressant administration on negative affective 72, 175–181. bias in depressed patients. Am. J. Psychiatry 166, 1178–1184. Field, T., 2010. Postpartum depression effects on early Hein, G., Singer, T., 2008. I feel how you feel but not always: the interactions, parenting, and safety practices: a review. Infant empathic brain and its modulation. Curr. Opin. Neurobiol. 18, Behav. Dev. 33, 1–6. 153–158. 98 brain research 1580 (2014) 78–101

Heinrichs, M., Baumgartner, T., Kirschbaum, C., Ehlert, U., 2003. associated with the production and care of offspring. Social support and oxytocin interact to suppress cortisol and J. Neuroendocrinol. 20, 515–525. subjective responses to psychosocial stress. Biol. Psychiatry Kober, H., Barrett, L.F., Joseph, J., Bliss-Moreau, E., Lindquist, K., 54, 1389–1398. Wager, T.D., 2008. Functional grouping and cortical– Herz, A., 1997. Endogenous opioid systems and alcohol addiction. subcortical interactions in emotion: a meta-analysis of Psychopharmacology 129, 99–111. neuroimaging studies. Neuroimage 42, 998–1031. Higham, J.P., Barr, C.S., Hoffman, C.L., Mandalaywala, T.M., Parker, Kochanska, G., 2002. Mutually responsive orientation between K.J., Maestripieri, D., 2011. Mu-opioid receptor (OPRM1) mothers and their young children: a context for the early variation, oxytocin levels and maternal attachment in free- development of conscience. Curr. Dir. Psychol. Sci. 11, ranging rhesus macaques Macaca mulatta. Behav. Neurosci. 191–195. 125, 131–136. Kosfeld, M., Heinrichs, M., Zak, P.J., Fischbacher, U., Fehr, E., 2005. Hrdy, S.B., 2000. In: Mother Nature: Maternal Instincts and How Oxytocin increases trust in humans. Nature 435, 673–676. They Shape the Human Species. Ballantine Books. Kringelbach, M.L., Lehtonen, A., Squire, S., Harvey, A.G., Craske, Iacoboni, M., 2009. Neurobiology of imitation. Curr. Opin. M.G., Holliday, I.E., Green, A.L., Aziz, T.Z., Hansen, P.C., Neurobiol. 19, 661–665. Cornelissen, P.L., Stein, A., 2008. A specific and rapid neural Insel, T.R., 1997. A neurobiological basis of social attachment. signature for parental instinct. PLoS One 3, e1664. Am. J. Psychiatry 154, 726–735. Kuo, P.X., Carp, J., Light, K.C., Grewen, K.M., 2012. Neural Insel, T.R., Young, L.J., 2001. The neurobiology of attachment. responses to infants linked with behavioral interactions and Nat. Rev. Neurosci. 2, 129–136. testosterone in fathers. Biol. Psychology 91, 302–306. Ishak, W.W., Kahloon, M., Fakhry, H., 2011. Oxytocin role in Lahey, B.B., Michalska, K.J., Liu, C., Chen, Q., Hipwell, A.E., enhancing well-being: a literature review. J. Affect. Disord. 130, Chronis-Tuscano, A., Waldman, I.D., Decety, J., 2012. 1–9. Preliminary genetic imaging study of the association between Jaffari-Bimmel, N., Juffer, F., van IJzendoorn, M.H., Bakermans- estrogen receptor-alpha gene polymorphisms and harsh Kranenburg, M.J., Mooijaart, A., 2006. Social development from human maternal parenting. Neurosci. Lett. 525, 17–22. infancy to adolescence: longitudinal and concurrent factors in Lamb, M.E., 1977. A re-examination of the infant social world. an adoption sample. Dev. Psychol. 42, 1143–1153. Hum. Dev. 20, 65–85. Joosen, K.J., Mesman, J., Bakermans-Kranenburg, M.J., Lamb, M.E., 2004. In: The Role of the Father in Child Development van IJzendoorn, M.H., 2012. Maternal sensitivity to infants in 4th ed. Wiley and Sons, New Jersey. Landi, N., Montoya, J., Kober, H., Rutherford, H.J., Mencl, W.E., various settings predicts harsh discipline in toddlerhood. Worhunsky, P.D., Potenza, M.N., Mayes, L.C., 2011. Maternal Attach. Hum. Dev. 14, 101–117. neural responses to infant cries and faces: relationships with Kaffman, A., Meaney, M.J., 2007. Neurodevelopmental sequelae of substance use. Front. Psychiatry 2, 32. postnatal maternal care in rodents: clinical and research Laurent, H.K., Ablow, J.C., 2012. A cry in the dark: depressed implications of molecular insights. J. Child Psychol. Psychiatry mothers show reduced neural activation to their own infant’s 48, 224–244. cry. Soc. Cogn. Affect. Neurosci. 7, 125–134. Kagerbauer, S.M., Martin, J., Schuster, T., Blobner, M., Kochs, E.F., Laurent, H.K., Ablow, J.C., 2013. A face a mother could love: Landgraf, R., 2013. Plasma oxytocin and vasopressin do not depression-related maternal neural responses to infant predict neuropeptide concentrations in human cerebrospinal emotion faces. Soc. Neurosci. 8, 228–239. fluid. J. Neuroendocrinol. 25, 668–673. Laurent, H.K., Stevens, A., Ablow, J.C., 2011. Neural correlates of Keller, H., Lohaus, A., Vo¨ lker, S., Elben, C., Ball, J., 2003. Warmth hypothalamic–pituitary–adrenal regulation of mothers with and contingency and their relationship to maternal attitudes their infants. Biol. Psychiatry 70, 826–832. toward parenting. J. Genet. Psychol. 164, 275–292. Leckman, J.F., Feldman, R., Swain, J.E., Eicher, V., Thompson, N., Keren, M., Feldman, R., Eidelman, A.I., Sirota, L., Lester, B., 2003. Mayes, L.C., 2004. Primary parental preoccupation: circuits, Clinical interview ofr high-risk parents of premature infants genes, and the crucial role of the environment. J. Neural (CLIP): relations to mother–infant interaction. Infant Ment. Transm. 111, 753–771. Health J. 24, 93–110. Leckman, J.F., Mayes, L.C., 1999. Preoccupations and behaviors Kim, P., Feldman, R., Mayes, L.C., Eicher, V., Thompson, N., associated with romantic and parental love. Perspectives on Leckman, J.F., Swain, J.E., 2011. Breastfeeding, brain activation the origin of obsessive–compulsive disorder. Child Adolesc. to own infant cry, and maternal sensitivity. J. Child Psychol. Psychiatr. Clin. N. Am. 8, 635–665. Psychiatry 52, 907–915. Leckman, J.F., Mayes, L.C., Feldman, R., Evans, D.W., King, R.A., Kim, P., Leckman, J.F., Mayes, L.C., Feldman, R., Wang, X., Swain, J. Cohen, D.J., 1999. Early parental preoccupations and behaviors E., 2010a. The plasticity of human maternal brain: and their possible relationship to the symptoms of obsessive– longitudinal changes in brain anatomy during the early compulsive disorder. Acta Psychiatr. Scand. 396 (Suppl.), postpartum period. Behav. Neurosci. 124, 695–700. S1–S26. Kim, P., Leckman, J.F., Mayes, L.C., Newman, M.A., Feldman, R., LeDoux, J., 2003. The emotional brain, fear, and the amygdala. Swain, J.E., 2010b. Perceived quality of maternal care in Cell. Mol. Neurobiol. 23, 727–738. childhood and structure and function of mothers’ brain. Dev. Leerkes, E.M., 2011. Maternal sensitivity during distressing tasks: Sci. 13, 662–673. a unique predictor of attachment security. Infant Behav. Dev. Kim, P., Mayes, L., Feldman, R., Leckman, J.F., Swain, J.E., 2013. 34, 443–446. Early postpartum parental preoccupation and positive Leerkes, E.M., Nayena Blankson, A., O’Brien, M., 2009. Differential parenting thoughts: relationship with parentinfant effects of maternal sensitivity to infant distress and interaction. Infant Ment. Health J. 34, 104–116. nondistress on social–emotional functioning. Child Dev. 80, Kim, P., Rigo, P., Mayes, L.C., Feldman, R., Leckman, J.F., Swain, J.E., 762–775. 2014. Neural plasticity in human fathers’ brains: associations Leibenluft, E., Gobbini, M.I., Harrison, T., Haxby, J.V., 2004. with depressive moods and parenting. Soc. Neurosci. (under Mothers’ neural activation in response to pictures of their review). children and other children. Biol. Psychiatry 56, 225–232. Kinsley, C.H., Lambert, K.G., 2008. Reproduction-induced Lenzi, D., Trentini, C., Pantano, P., Macaluso, E., Iacoboni, M., neuroplasticity: natural behavioural and neuronal alterations Lenzi, G.L., Ammaniti, M., 2009. Neural basis of maternal brain research 1580 (2014) 78–101 99

communication and emotional expression processing during Montoya, J.L., Landi, N., Kober, H., Worhunsky, P.D., infant preverbal stage. Cereb. Cortex 19, 1124–1133. Rutherford, H.J.V., Mencl, W.E., Mayes, L.C., Potenza, M.N., Levine, A., Zagoory-Sharon, O., Feldman, R., Weller, A., 2007. 2012. Regional brain responses in nulliparous women to Oxytocin during pregnancy and early postpartum: individual emotional infant stimuli. PLoS One 7, e36270. patterns and maternal-fetal attachment. Peptides 28, Morris, J.P., Pelphrey, K.A., McCarthy, G., 2005. Regional brain 1162–1169. activation evoked when approaching a virtual human on a Light, K.C., Smith, T.E., Johns, J.M., Brownley, K.A., Hofheimer, J.A., virtual walk. J. Cogn. Neurosci. 17, 1744–1752. Amico, J.A., 2000. Oxytocin responsivity in mothers of infants: Moser, D.A., Aue, T., Wang, Z., Rusconi Serpa, S., Favez, N., a preliminary study of relationships with blood pressure Peterson, B.S., Schechter, D.S., 2013. Limbic brain responses in during laboratory stress and normal ambulatory activity. mothers with post-traumatic stress disorder and comorbid Health Psychol. 19, 560–567. dissociation to video clips of their children. Stress 16, 493–502. Lorberbaum, J.P., Newman, J.D., Dubno, J.R., Horwitz, A.R., Nahas, Moses-Kolko, E.L., Hipwell, A., Swain, J.E., 2014. In search of Z., Teneback, C.C., Bloomer, C.W., Bohning, D.E., Vincent, D., neural endophenotypes of postpartum psychopathology and Johnson, M.R., Emmanuel, N., Brawman-Mintzer, O., Book, S. disrupted maternal caregiving. J. Neuroendocrinol. (in press). W., Lydiard, R.B., Ballenger, J.C., George, M.S., 1999. Feasibility Moses-Kolko, E.L., Perlman, S.B., Wisner, K.L., James, J., Saul, A.T., of using fMRI to study mothers responding to infant cries. Phillips, M.L., 2010. Abnormally reduced dorsomedial Depress. Anxiety 10, 99–104. prefrontal cortical activity and effective connectivity with Lorberbaum, J.P., Newman, J.D., Horwitz, A.R., Dubno, J.R., Lydiard, amygdala in response to negative emotional faces in R.B., Hamner, M.B., Bohning, D.E., George, M.S., 2002. A postpartum depression. Am. J. Psychiatry 167, 1373–1380. potential role for thalamocingulate circuitry in human Murray, L., Fiori-Cowley, A., Hooper, R., Cooper, P., 1996. The impact of postnatal depression and associated adversity on maternal behavior. Biol. Psychiatry 51, 431–445. Lupien, S.J., McEwen, B.S., Gunnar, M.R., Heim, C., 2009. Effects of early mother–infant interactions and later infant outcome. Child Dev. 67, 2512–2526. stress throughout the lifespan on the brain, behaviour and Musser, E.D., Kaiser-Laurent, H., Ablow, J.C., 2012. The neural cognition. Nat. Rev. Neurosci. 10, 434–445. correlates of maternal sensitivity: an fMRI study. Dev. Cogn. MacDonald, K., 1987. Parent–child physical play with rejected, Neurosci. 2, 428–436. neglected, and popular boys. Dev. Psychol. 23, 705–711. Neumann, I.D., 2008. Brain oxytocin: a key regulator of emotional MacDonald, K., 1992. Warmth as a developmental construct: an and social behaviours in both females and males. J. evolutionary analysis. Child Dev. 63, 753–773. Neuroendocrinol. 20, 858–865. Macdonald, K., Feifel, D., 2013. Helping oxytocin deliver: Nishitani, S., Doi, H., Koyama, A., Shinohara, K., 2011. Differential considerations in the development of oxytocin-based prefrontal response to infant facial emotions in mothers therapeutics for brain disorders. Front. Neurosci. 7, 35. compared with non-mothers. Neurosci. Res. 70, 183–188. Macdonald, K., Macdonald, T.M., Brune, M., Lamb, K., Nitschke, J.B., Nelson, E.E., Rusch, B.D., Fox, A.S., Oakes, T.R., Wilson, M.P., Golshan, S., Feifel, D., 2013. Oxytocin and Davidson, R.J., 2004. Orbitofrontal cortex tracks positive mood psychotherapy: a pilot study of its physiological, behavioral in mothers viewing pictures of their newborn infants. and subjective effects in males with depression. Neuroimage 21, 583–592. Psychoneuroendocrinology. Noriuchi, M., Kikuchi, Y., Senoo, A., 2008. The functional MacLean, P.D., 1990. In: The Triune Brain in Evolution: Role in neuroanatomy of maternal love: mother’s response to infant’s Paleocerebral Functions. Plenum Press, New York. attachment behaviors. Biol. Psychiatry 63, 415–423. Marazziti, D., Dell’osso, B., Baroni, S., Mungai, F., Catena, M., Nowak, R., Keller, M., Levy, F., 2011. Mother–young relationships Rucci, P., Albanese, F., Giannaccini, G., Betti, L., Fabbrini, L., in sheep: a model for a multidisciplinary approach of the Italiani, P., Del Debbio, A., Lucacchini, A., Dell’osso, L., 2006. study of attachment in mammals. J. Neuroendocrinol. 23, A relationship between oxytocin and anxiety of romantic 1042–1053. attachment. Clin. Pract. Epidemol. Ment. Health 2, 28. Numan, M., Insel, T.R., 2003a. In: The Neurobiology of Parental Mascaro, J.S., Hackett, P.D., Gouzoules, H., Lori, A., Rilling, J.K., Behavior. Springer-Verlag, New York. 2013a. Behavioral and genetic correlates of the neural Numan, M., Insel, T.R., 2003b. In: The Neurobiology of Parental response to infant crying among human fathers. Soc. Cogn. Behavior. Springer, New York. Affect. Neurosci.. Numan, M., Woodside, B., 2010. Maternity: neural mechanisms, Mascaro, J.S., Hackett, P.D., Rilling, J.K., 2013b. Testicular volume motivational processes, and physiological adaptations. Behav. is inversely correlated with nurturing-related brain activity in Neurosci. 124, 715–741. human fathers. Proc. Natl. Acad. Sci. USA 110, 15746–15751. Nystrom, K., Ohrling, K., 2004. Parenthood experiences during the McElwain, N.L., Booth-Laforce, C., 2006. Maternal sensitivity to child’s first year: literature review. J. Adv. Nurs. 46, 319–330. infant distress and nondistress as predictors of infant–mother Olff, M., Frijling, J.L., Kubzansky, L.D., Bradley, B., Ellenbogen, M. attachment security. J. Fam. Psychol. 20, 247–255. A., Cardoso, C., Bartz, J.A., Yee, J.R., van Zuiden, M., 2013. The McEwen, B.S., 2008. Understanding the potency of stressful early role of oxytocin in social bonding, stress regulation and life experiences on brain and body function. 57 mental health: an update on the moderating effects of context (Suppl. 2), S11–S15. and interindividual differences. Psychoneuroendocrinology Mercer, R.T., 1985. The process of maternal role attainment over 38, 1883–1894. the first year. Nurs. Res. 34, 198–204. Pantazatos, S.P., Talati, A., Pavlidis, P., Hirsch, J., 2012. Decoding Milner, J.S., 1993. Social information-processing and physical unattended fearful faces with whole-brain correlations: an child-abuse. Clin. Psychol. Rev. 13, 275–294. approach to identify condition-dependent large-scale Milner, J.S., 2003. Social information processing in high-risk and functional connectivity. PLoS Comput. Biol. 8, e1002441. physically abusive parents. Child Abuse Negl. 27, 7–20. Papousek, H., Papousek, M., 2002. Intuitive parenting. In: Mitchell, J.P., 2009. Inferences about mental states. Philos. Trans. Bornstein, M.H. (Ed.), Handbook of Parenting. Erlbaum, R. Soc. Lond. B: Biol. Sci. 364, 1309–1316. Mahwah, N.J., pp. 183–203. Molitor, A., Mayes, L.C., 2010. Problematic dyadic interaction Paquette, D., 2004. Theorizing the father–child relationship: among toddlers and their polydrug–cocaine-using mothers. mechanisms and developmental outcomes. Hum. Dev. 47, Infant Ment. Health J. 31, 121–140. 193–219. 100 brain research 1580 (2014) 78–101

Parsons, C.E., Young, K.S., Kumari, N., Stein, A., Kringelbach, M.L., Seifritz, E., Esposito, F., Neuhoff, J.G., Luthi, A., Mustovic, H., 2011. The motivational salience of infant faces is similar for Dammann, G., von Bardeleben, U., Radue, E.W., Cirillo, S., men and women. PLoS One 6, e20632. Tedeschi, G., Di Salle, F., 2003. Differential sex-independent Parsons, C.E., Young, K.S., Mohseni, H., Woolrich, M.W., Thomsen, amygdala response to infant crying and laughing in parents K.R., Joensson, M., Murray, L., Goodacre, T., Stein, A., versus nonparents. Biol. Psychiatry 54, 1367–1375. Kringelbach, M.L., 2013. Minor structural abnormalities in the Shahrokh, D.K., Zhang, T.Y., Diorio, J., Gratton, A., Meaney, M.J., infant face disrupt neural processing: a unique window into 2010. Oxytocin–dopamine interactions mediate variations in early caregiving responses. Soc. Neurosci. 8, 268–274. maternal behavior in the rat. Endocrinology 151, 2276–2286. Pearson, R.M., Cooper, R.M., Penton-Voak, I.S., Lightman, S.L., Shaver, P., Schwartz, J., Kirson, D., O’Connor, C., 1987. Emotion Evans, J., 2010. Depressive symptoms in early pregnancy knowledge: further exploration of a prototype approach. J. disrupt attentional processing of infant emotion. Psychol. Personal. Soc. Psychol. 52, 1061–1086. Med. 40, 621–631. Simmons, D., Self, D.W., 2009. Role of mu- and delta-opioid Pearson, R.M., Lightman, S.L., Evans, J., 2011. The impact of receptors in the nucleus accumbens in cocaine-seeking breastfeeding on mothers’ attentional sensitivity towards behavior. Neuropsychopharmacology 34, 1946–1957. infant distress. Infant Behav. Dev. 34, 200–205. Skrundz, M., Bolten, M., Nast, I., Hellhammer, D.H., Pearson, R.M., Lightman, S.L., Evans, J., 2012. Symptoms of Meinlschmidt, G., 2011. Plasma oxytocin concentration during depression during pregnancy are associated with increased pregnancy is associated with development of postpartum systolic blood pressure responses towards infant distress. depression. Neuropsychopharmacology 36, 1886–1893. Arch. Womens Ment. Health 15, 95–105. Soltis, J., 2004. The signal functions of early infant crying. Behav. Pedersen, C.A., Prange Jr., A.J., 1979. Induction of maternal Brain Sci. 27, 443–458 (discussion 459–490). behavior in virgin rats after intracerebroventricular Sripada, C., Angstadt, M., Kessler, D., Phan, K.L., Liberzon, I., administration of oxytocin. Proc. Natl. Acad. Sci. USA 76, Evans, G.W., Welsh, R.C., Kim, P., Swain, J.E., 2013. Volitional regulation of emotions produces distributed alterations in 6661–6665. Pruett, K.D., 1998. Role of the father. Pediatrics 102, 1253–1261. connectivity between visual, attention control, and default Quirin, M., Kuhl, J., Dusing, R., 2011. Oxytocin buffers cortisol networks. Neuroimage 89C, 110–121. Strathearn, L., Fonagy, P., Amico, J., Montague, P.R., 2009. Adult responses to stress in individuals with impaired emotion attachment predicts maternal brain and oxytocin response to regulation abilities. Psychoneuroendocrinology 36, 898–904. infant cues. Neuropsychopharmacology 34, 2655–2666. Ramchandani, P., Psychogiou, L., 2009. Paternal psychiatric Strathearn, L., Li, J., Fonagy, P., Montague, P.R., 2008. What’s in a disorders and children’s psychosocial development. Lancet smile? Maternal brain responses to infant facial cues. 374, 646–653. Pediatrics 122, 40–51. Ranote, S., Elliott, R., Abel, K.M., Mitchell, R., Deakin, J.F., Appleby, Strathearn, L., Li, J., Montague, P.R., 2005. An fMRI study of L., 2004. The neural basis of maternal responsiveness to maternal mentalization: having the baby’s mind in mind. infants: an fMRI study. Neuroreport 15, 1825–1829. Neuroimage 26 (S25). Riem, M.M., Bakermans-Kranenburg, M.J., Pieper, S., Tops, M., Striepens, N., Scheele, D., Kendrick, K.M., Becker, B., Schafer, L., Boksem, M.A., Vermeiren, R.R., van Ijzendoorn, M.H., Schwalba, K., Reul, J., Maier, W., Hurlemann, R., 2012. Oxytocin Rombouts, S.A., 2011. Oxytocin modulates amygdala, insula, facilitates protective responses to aversive social stimuli in and inferior frontal gyrus responses to infant crying: a males. Proc. Natl. Acad. Sci. USA 109, 18144–18149. randomized controlled trial. Biol. Psychiatry 70, 291–297. Swain, J.E., 2011a. Brain Imaging of Human Parent–Infant Riem, M.M., Bakermans-Kranenburg, M.J., van Ijzendoorn, M.H., Relationships. Arch. Women’s Ment. Health 14, s93–s94. Out, D., Rombouts, S., 2012. Attachment in the brain: adult Swain, J.E., 2011b. The human parental brain: in vivo attachment representations predict amygdala and behavioral neuroimaging. Prog. Neuropsychopharmacol. Biol. Psychiatry responses to infant crying. Attach. Hum. Dev. 14, 533–551. 35, 1242–1254. Rolls, E.T., 2004. The functions of the orbitofrontal cortex. Brain Swain, J.E., Kim, P., Ho, S.S., 2011. Neuroendocrinology of parental Cogn. 55, 11–29. response to baby-cry. J. Neuroendocrinol. 23, 1036–1041. Rosenblatt, J.S., Ceus, K., 1998. Estrogen implants in the medial Swain, J.E., Konrath, S., Brown, S.L., Finegood, E.D., Akce, L.B., preoptic area stimulate maternal behavior in male rats. Horm. Dayton, C.J., Ho, S.S., 2012. Parenting and beyond: common Behav. 33, 23–30. neurocircuits underlying parental and altruistic caregiving. Ross, H.E., Young, L.J., 2009. Oxytocin and the neural mechanisms Parent., Sci. Pract. 12, 115–123. regulating social cognition and affiliative behavior. Front. Swain, J.E., Leckman, J.F., Mayes, L.C., Feldman, R., Constable, R.T., Neuroendocrinol. 30, 534–547. Schultz, R.T., 2003. The neural circuitry of parent–infant Rutter, M., 2013. Annual research review: resilience – clinical attachment in the early postpartum. Am. Coll. implications. J. Child Psychol. Psychiatry 54, 474–487. Neuropsychopharmacol. Saxe, R., 2006. Uniquely human social cognition. Curr. Opin. Swain, J.E., Leckman, J.F., Mayes, L.C., Feldman, R., Constable, R.T., Neurobiol. 16, 235–239. Schultz, R.T., 2004a. Neural substrates of human parent–infant Schechter, D.S., Moser, D.A., Wang, Z., Marsh, R., Hao, X., Duan, Y., attachment in the postpartum. Biol. Psychiatry 55, 153s. Yu, S., Gunter, B., Murphy, D., McCaw, J., Kangarlu, A., Swain, J.E., Leckman, J.F., Mayes, L.C., Feldman, R., Hoyt, E., Kang, Willheim, E., Myers, M.M., Hofer, M.A., Peterson, B.S., 2012. An H., Kim, P., David, D., Nguyen, S., Constable, R.T., Schultz, R.T., fMRI study of the brain responses of traumatized mothers to 2014. Functional brain activations of parents listening to their viewing their toddlers during separation and play. Soc. Cogn. own baby-cry change over the early postpartum. Dev. Affect. Neurosci. 7, 969–979. Psychobiol. (under review). Schilbach, L., Timmermans, B., Reddy, V., Costall, A., Bente, G., Swain, J.E., Leckman, J.F., Mayes, L.C., Feldman, R., Schultz, R.T., Schlicht, T., Vogeley, K., 2013. Toward a second-person 2005. Early human parent-infant bond development: fMRI, neuroscience. Behav. Brain Sci. 36, 393–414. thoughts and behaviors. Biol. Psychiatry 57, 112s. Schuengel, C., Bakermans-Kranenburg, M.J., Van, I.M.H., 1999. Swain, J.E., Leckman, J.F., Mayes, L.C., Feldman, R., Schultz, R.T., Frightening maternal behavior linking unresolved loss and 2006. Own baby pictures induce parental brain activations disorganized infant attachment. J. Consult. Clin. Psychol. 67, according to psychology, experience and postpartum timing. 54–63. Biol. Psychiatry 59, 126s. brain research 1580 (2014) 78–101 101

Swain, J.E., Lorberbaum, J.P., 2008. Imaging the human parental infants with autistic disorder and typical development: an brain. Neurobiol. Parent. Brain, 83–100. fMRI study. Res. Dev. Disabil. 33, 2255–2264. Swain, J.E., Lorberbaum, J.P., Kose, S., Strathearn, L., 2007. Brain Victor, T.A., Furey, M.L., Fromm, S.J., Bellgowan, P.S., Ohman, A., basis of early parent–infant interactions: psychology, Drevets, W.C., 2012. The extended functional neuroanatomy physiology, and in vivo functional neuroimaging studies. J. of emotional processing biases for masked faces in major Child Psychol. Psychiatry 48, 262–287. depressive disorder. PLoS One 7, e46439. Swain, J.E., Mayes, L.C., Leckman, J.F., 2004b. The development of Vigil, J.M., 2009. A socio-relational framework of sex differences parent–infant attachment through dynamic and interactive in the expression of emotion. Behav. Brain Sci. 32, 375–390 signaling loops of care and cry. Behav. Brain Sci. 27, 472–473. (discussion 391–428). Swain, J.E., Tasgin, E., Mayes, L.C., Feldman, R., Constable, R.T., Volling, B.L., McElwain, N.L., Notaro, P.C., Herrera, C., 2002. Leckman, J.F., 2008. Maternal brain response to own baby-cry Parents’ emotional availability and infant emotional is affected by cesarean section delivery. J. Child Psychol. competence: predictors of parent–infant attachment and Psychiatry 49, 1042–1052. emerging self-regulation. J. Fam. Psychol. 16, 447–465. Tabak, B.A., McCullough, M.E., Szeto, A., Mendez, A.J., McCabe, P. Voorthuis, A., Riem, M.M., Van Ijzendoorn, M.H., Bakermans- M., 2011. Oxytocin indexes relational distress following Kranenburg, M.J., 2014. Reading the mind in the infant eyes: interpersonal harms in women. Psychoneuroendocrinology Paradoxical effects of oxytocin on neural activity and emotion 36, 115–122. recognition in watching pictures of infant faces. Brain Res. Tamis-LeMonda, C.S., Bornstein, M.H., Baumwell, L., 2001. 2014, epub ahead of print 1580, 151–159. Maternal responsiveness and children’s achievement of Wan, M.W., Abel, K.M., Green, J., 2008a. The transmission of risk language milestones. Child Dev. 72, 748–767. to children from mothers with schizophrenia: a Taylor, S.E., Gonzaga, G.C., Klein, L.C., Hu, P., Greendale, G.A., developmental psychopathology model. Clin. Psychol. Rev. 28, Seeman, T.E., 2006. Relation of oxytocin to psychological 613–637. stress responses and hypothalamic–pituitary–adrenocortical Wan, M.W., Downey, D., Strachan, H., Elliott, R., Williams, S.R., axis activity in older women. Psychosom. Med. 68, 238–245. Abel, K.M., 2014. The neural basis of maternal bonding. PLoS Taylor, S.E., Seeman, T.E., Eisenberger, N.I., Kozanian, T.A., Moore, One 9, e88436. A.N., Moons, W.G., 2010. Effects of a supportive or an Wan, M.W., Green, J., Elsabbagh, M., Johnson, M., Charman, T., unsupportive audience on biological and psychological Plummer, F., Team, B., 2013. Quality of interaction between at- responses to stress. J. Personal. Soc. Psychol. 98, 47–56. risk infants and caregiver at 12–15months is associated with Theodoridou, A., Penton-Voak, I.S., Rowe, A.C., 2013. A direct 3-year autism outcome. J. Child Psychol. Psychiatry 54, examination of the effect of intranasal administration of 763–771. oxytocin on approach–avoidance motor responses to Wan, M.W., Penketh, V., Salmon, M.P., Abel, K.M., 2008b. Content emotional stimuli. PLoS One 8, e58113. and style of speech from mothers with schizophrenia towards Thompson-Booth, C., Viding, E., Mayes, L.C., Rutherford, H.J., their infants. Psychiatry Res. 159, 109–114. Hodsoll, S., McCrory, E.J., 2014. Here’s looking at you, kid: Wan, M.W., Warren, K., Salmon, M.P., Abel, K.M., 2008c. Patterns attention to infant emotional faces in mothers and non- of maternal responding in postpartum mothers with mothers. Dev. Sci. 17, 35–46. schizophrenia. Infant Behav. Dev. 31, 532–538. Uddin, L.Q., Iacoboni, M., Lange, C., Keenan, J.P., 2007. The self Weisman, O., Zagoory-Sharon, O., Feldman, R., 2012. Oxytocin and social cognition: the role of cortical midline structures administration to parent enhances infant physiological and and mirror neurons. Trends Cogn. Sci. 11, 153–157. behavioral readiness for social engagement. Biol. Psychiatry van Leengoed, E., Kerker, E., Swanson, H.H., 1987. Inhibition of 72, 982–989. post-partum maternal behaviour in the rat by injecting an Winnicott, D.W., 1956. Primary maternal preoccupation. In: oxytocin antagonist into the cerebral ventricles. J. Endocrinol. Winnicott, D.W. (Ed.), Through Paediatrics to Psycho-analysis. 112, 275–282. Hogarth, London. Van Zeijl, J., Mesman, J., Van IJzendoorn, M.H., Bakermans- Winslow, J.T., Hastings, N., Carter, C.S., Harbaugh, C.R., Insel, T.R., Kranenburg, M.J., Juffer, F., Stolk, M.N., Koot, H.M., Alink, L.R. 1993. A role for central vasopressin in pair bonding in A., 2006. Attachment-based intervention for enhancing monogamous prairie voles. Nature 365, 545–548. sensitive discipline in mothers of 1-to 3-year-old children at Wittfoth-Schardt, D., Gru¨ nding, J., Wittfoth, M., Lanfermann, H., risk for externalizing behavior problems: a randomized Heinrichs, M., Domes, G., Buchheim, A., Gu¨ ndel, H., Waller, C., controlled trial. J. Consult. Clin. Psychol. 74, 994–1005. 2012. Oxytocin modulates neural reactivity to children’s faces Veenema, A.H., 2009. Early life stress, the development of as a function of social salience. Neuropsychopharmacology aggression and neuroendocrine and neurobiological 37, 1799–1807. correlates: what can we learn from animal models?. Front. Yogman, M.W., 1981. Games fathers and mothers play with their Neuroendocrinol. 30, 497–518. infants. Infant Ment. Health J. 2, 241–248. Venuti, P., Caria, A., Esposito, G., De Pisapia, N., Bornstein, M.H., Zak, P.J., Stanton, A.A., Ahmadi, S., 2007. Oxytocin increases de Falco, S., 2012. Differential brain responses to cries of generosity in humans. PLoS One 2, e1128. brain research 1580 (2014) 143–150

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Does oxytocin modulate variation in maternal caregiving in healthy new mothers?

Alya Elmadiha, Ming Wai Wana, Michael Numanb, Rebecca Elliottc, Darragh Downeyc, Kathryn M Abela,n aInstitute of Brain, Behaviour and Mental Health, University of Manchester, Manchester M13 9PL, UK bMichael Numan, Rio Rancho, NM 87144, United States cNeuroscience and Psychiatry Unit, Institute of Brain, Behaviour and Mental Health, University of Manchester, Manchester M13 9PL, UK article info abstract

Article history: Background: Maternal sensitivity to infant cues and developmental needs may be pivotal Accepted 15 January 2014 for social and cognitive development. Animal and recent human studies emphasise a Available online 23 January 2014 major role for Oxytocin (OT) in mediating sensitive caregiving but no study has examined Keywords: the relationship between OT and extreme variation in human maternal sensitivity. Plasma oxytocin Methods: From 105 expectant mothers, 80 underwent blind-rating of maternal sensitivity Natural variation at 4–6 months postpartum through free-play interaction with their infants. At 7–9 months Maternal sensitivity postpartum, 30 mothers at extremes of maternal sensitivity: 15 ‘sensitive mothers’ (high sensitivity mothers – HSMs, mean 4.47; SD 0.74) and 15 ‘less sensitive mothers’ (low Mother–infant interaction ¼ ¼ sensitivity mothers – LSMs, mean 2.13; SD 0.52) underwent plasma OT measurements Stress ¼ ¼ before and after 10 min infant play. Results: Baseline and post-interaction plasma OT was higher in LSMs than HSMs [F(1, 26) 8.42; p 0.01]. HSMs showed a trend towards ¼ ¼ significant reduction in plasma OT [t(14) 2.01; p 0.06] following play-interaction; no ¼ ¼ change was shown by LSMs [t(13) 0.14; p 0.89]. Conclusion Higher baseline OT levels in ¼À ¼ healthy LSMs may imply greater stress responses to the demands of caring for an infant, or past deficiencies in own parenting relationship and act as a biomarker for poor parental sensitivity. OT may be acting to reduce stress and anxiety in LSMs consistent with studies of plasma OT and stress in women. By contrast, in HSMs, play interaction with their infants maybe relaxing as indicated by significant reduction in plasma OT from baseline. Ascertainment of mothers in well-defined sensitivity groups might facilitate examination of distinct coping strategies in parents and better understanding of variation in parental caregiving behaviour and its potential for modulation by OT. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2014 Elsevier B.V. All rights reserved.

nCorrespondence to: Institute of Brain, Behaviour and Mental Health, University of Manchester, 3rd Floor East Jean McFarlane Building, Manchester M13 9PL, UK. E-mail addresses: [email protected] (A. Elmadih), [email protected] (M. Numan).

0006-8993/$ - see front matter & 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.brainres.2014.01.020 144 brain research 1580 (2014) 143–150

study considered a mothers' social relationships, including 1. Introduction relationships with their own parents in the context of examining maternal affiliative bond with own infants Human infants are immature creatures depending wholly on (Feldman et al., 2011). In the present study, we set out to their caregivers for survival. Maternal caregiving is a complex examine (i) baseline and (ii) stimulated plasma OT responses individualised behaviour (see Barrett and Fleming 2011), an in women occupying opposite ends of the spectrum of essential element of which is being soothed at stressful times natural variations in maternal sensitivity; and (iii) to (Feldman et al., 2007). Recent animal (Champagne et al., 2001; determine whether measurement of OT could distinguish Francis et al., 2000; Numan and Stolzenberg, 2009; Ross and between sensitivity groups. Finally, we explored the Young, 2009) and human studies suggest that the nonapep- relationship between a mother's baseline and infant- tide oxytocin (OT) is essential to enhance social competence challenge plasma OT levels, and her own parenting experi- and to initiate and promote adequate maternal caregiving ence. Based on the studies which emphasise a positive behaviour (Feldman et al., 2010a, b; Gordon et al., 2010; Lee relationship between OT and parenting, we hypothesised et al., 2009; Numan and Woodside, 2010; Swain et al., 2008). that: (1) higher sensitivity mothers (HSMs) would show Mothers who showed a rising pattern of plasma OT across significantly higher plasma OT levels at baseline and follow- pregnancy and the early postpartum also reported stronger ing interaction with their infants than lower sensitivity maternal-fetal attachment (Levine et al., 2007). Similarly, mothers (LSMs); (2) post-interaction plasma OT in each group fi higher plasma OT levels during pregnancy and at the rst of mothers (HSMs/LSMs) would be positively related to a postpartum month were correlated with higher levels of favourable rating of the quality of her own parenting maternal postpartum behaviours, such as gaze, vocalisations experience. and positive affect in new mothers (Feldman et al., 2007). Maternal synchrony, defined as ‘episodes when mother and infant coordinate their positive social engagement’ (Atzil 2. Results et al., 2011), was also found to be correlated with maternal plasma OT levels, while maternal intrusiveness, ‘inappropri- Among the 30 mothers included in the current study: 15 were ate response from mother’, was not. classified as ‘sensitive’ mothers (blind rated as 4–7 on the OT has also been examined in relation to the mother's MACI sensitivity scale; mean 4.47; SD 0.74) and 15 mothers ¼ ¼ own attachment experience. Higher levels of plasma OT have classified as ‘less sensitive’ mothers (blind rated as 1–3 on the been reported following mother–infant physical interaction MACI sensitivity scale; mean 2.13; SD 0.52). For description ¼ ¼ among mothers reporting secure attachment (with own purposes, the two sub-groups are referred to here as ‘high parents) compared with those reporting insecure attachment sensitivity mothers’ (HSMs) and ‘low sensitivity mothers’ (Strathearn et al., 2009). Even among non-parents, plasma (LSMs), respectively. OT levels have been positively correlated with self-reported recall of parental care (maternal and paternal care) (Feldman 2.1. Assessing for confounders et al., 2012; Gordon et al., 2008). Similarly, among children, urinary OT in 4.5 year olds who were raised by their own One outlier was excluded from analyses (scored43 SD in all parents showed a trend for higher levels compared with OT assessments). Household income was higher in HSMs children raised in orphanages (Fries et al., 2005). Although compared with LSMs and accordingly was adjusted for in the most of the studies that link plasma OT to own parental subsequent analysis. The two groups did not differ in mater- experience involved a one-off plasma OT assessment (e.g. nal age, marital status, maternal education (in years), parity, Fries et al., 2005; Gordon et al., 2008), post-infant challenge OT infant birthweight, infant gender, mode of delivery, mode of levels (rather than baseline levels) were found to be more feeding, or postpartum stage (Table 1). No significant differ- relevant in reflecting the quality of experiences of being ences between high and low sensitivity mothers were found parented (Strathearn et al., 2009). in anxiety scores (F (1, 27) 2.08, p 0.16) nor in depression ¼ ¼ Recent evidence from studies in women generally (not scores (F (1, 27) 0.46, p 0.50). ¼ ¼ pregnant or new mothers) suggests a role for OT in the No relationship was found between current breastfeeding regulation of stress related to interpersonal relationships status and maternal sensitivity F (1, 27) 0.17, p 0.68. ¼ ¼ (Tabak et al., 2011). Such studies report higher levels of Also, breastfeeding mothers did not significantly differ from plasma OT associated with stress in social relationships, other mothers in either baseline plasma OT measurement including relationships with romantic partners (Feldman (mean 306.54 vs. 259.16, F (1, 27) 1.78, p 0.19) or post- ¼ ¼ ¼ et al., 2011; Marazziti et al., 2006; Taylor et al., 2010), own interaction measurement (mean 275.17 vs. 254.61, F (1, 27) ¼ ¼ mother (Taylor et al., 2006), or best friend (Taylor et al., 2006). 0.30, p 0.59). ¼ This is consistent with the well-recognised anti-anxiety/ anxiolytic and anti-stress effect of OT (Guzman et al., 2013; 2.2. Plasma oxytocin Numan and Woodside, 2010) and suggests OT maybe released in stressful situations in order to decrease or moderate stress Three samples were collected from each participant (Table 2). responses (Marazziti et al., 2006). A high level of OT individual stability was found in HSMs To date, parenting studies that focus on OT in relation to (r 0.79–0.96; po0.01) and LSMs (r 0.92–0.99; po0.01). ¼ ¼ maternal affiliative behaviour have not included rigorous Repeated measure ANOVA showed no overall change in examination of maternal sensitivity. In addition, only one the mean level of OT from baseline to post-interaction brain research 1580 (2014) 143–150 145

Table 1 – Comparing demographic and obstetric characteristics of mothers grouped by level of maternal sensitivity, excluding one outlier (high sensitivity mothers HSMs, N 15 and low sensitivity ¼ mothers LSMs, N 14). ¼ Characteristic HSMs LSMs t (27) Chi-square p-value (N 15) (N 14) test ¼ ¼ Mean [SD] Maternal age (years) 30.40 [5.37] 27.64 [4.77] 1.46 0.16 À Average maternal education (years) 15.06 [2.82] 12.77 [2.76] 1.94 0.06 À Average annual household income 33.00 [4.61] 26.12 [4.13] 2.11 0.04 À (thousand pounds) Infant birthweight (kg) 3.44 [0.44] 3.20 [0.53] 1.10 0.28 À Postpartum stage (weeks) 35.93 [2.81] 34.64 [3.22] 1.32 0.20 Frequency (%) Married/cohabiting 13 (86.7) 11 (78.6) 0.56 0.65 Primiparous 6 (40.0) 8 (57.1) 1.65 0.21 Infant gender (female) 10 (66.7) 7 (50) 0.83 0.36 Mode of delivery (vaginal) 10 (66.7) 11 (78.6) 0.51 0.47 Mode of feeding (breast) 3 (20.0) 3 (20.0) 0.55n 0.64

n Fisher exact test.

Table 2 – Mean plasma OT levels (pg/ml) among the high (N 15) and low (N 14) sensitivity group of mothers. ¼ ¼ Oxytocin High sensitivity Low sensitivity sample mothers (N 15) mean mothers (N 14) mean ¼ ¼ [SD] [SD]

OT1 235.09 [83.51] 301.87 [39.15] OT2 223.67 [83.43] 303.27 [34.67] OT3 210.01 [81.58] 301.56 [38.12]

Note: OT1: Oxytocin measured before mother–infant interaction, OT2 and OT3: Oxytocin measured after mother–infant interaction.

(within subject effect) Greenhouse-Geisser [F (1.37, 35.51) Fig. 1 – Means and box plots for plasma oxytocin measured ¼ 1.54; p 0.23]. Next, we examined between subject effects before (OT1) and after mother–infant interaction (OT2 and ¼ [F (1, 26) 8.42; p 0.01] and found significant difference in OT OT3) among the high sensitivity mothers (HSMs, N 15) and ¼ ¼ ¼ through the three assessment points when the two groups of low sensitivity mothers (LSMs, N 14), controlling for ¼ sensitivity were compared (between subject effects). To confirm household income. Key: Dashed line represents the means the direction of significance, this was examined further by paired of the three OT assessments among LSMs, and solid line t-tests; this showed a trend for a significant drop in OT levels represents means of the three assessments among among high sensitivity mothers; from OT (OT1) (mean 235.09; the HSMs. pre ¼ SD 83.51) to OT ((OT2 OT3)/2) (mean 216.84; SD 79.18); ¼ post þ ¼ ¼ t (14) 2.01; p 0.06 compared with no significant difference ¼ ¼ between the two time of measurements among low sensitivity OTpre and the mean of OT2 and OT3 (OTpost) were con- mothers (mean OT 301.87; SD 0.39.15, and OT pre¼ ¼ post sidered to test the relationship with own parenting experi- mean 302. 15; SD 36.27, respectively); t (13) 0.14; p 0.89, ¼ ¼ ¼À ¼ ence. Own maternal overprotection was positively correlated indicating the moderate change in response to interaction is with both OT and OT among LSMs (r 0.62; p 0.02; and fi pre post ¼ ¼ con ned to the HSMs (Fig. 1). r 0.63; p 0.02, respectively), such that the greater the over- ¼ ¼ protection they perceived from their mother in family of 2.3. Relationship between plasma OT and mothers' own origin, the higher the baseline OT level and the less change in parenting experience OT following child play challenge. However, similar relation-

ship between OTpre and OTpost levels and own maternal Controlling for household income, the two groups did not differ overprotection was not found among HSMs (r 0.28; p 0.31; ¼ ¼ in own maternal care (transformed) [F (1, 26) 0.01; p 0.93], and r 0.07; p 0.79, respectively) (Fig. 2/Table 3). Own mater- ¼ ¼ ¼ ¼ own paternal care (transformed) [F (1, 26) 0.41; p 0.53], own nal care, own paternal care, or own paternal overprotection ¼ ¼ maternal overprotection [F (1, 26) 2.41; p 0.13], or own were not correlated with plasma OT levels in any of the ¼ ¼ paternal overprotection [F (1, 26) 0.52; p 0.48]. groups (all p's40.22). ¼ ¼ 146 brain research 1580 (2014) 143–150

LSMs. We also found that greater levels of baseline and post- 3. Discussion interaction plasma OT in LSMs were associated with less favourable reports of their own experience of being parented We sought to examine how opposite extremes of the dis- (i.e. higher own maternal overprotection). tribution in maternal sensitivity to infant cues are related to These findings contradict recent parenting studies report- maternal plasma OT levels in a community sample of healthy ing positive correlations between plasma OT levels and new mothers. Contrary to our main hypotheses, we found maternal (or paternal) behaviour (Atzil et al., 2011; Feldman that (1) HSMs have lower baseline plasma OT levels compared et al., 2007; 2010a, b; Gordon et al., 2010). In the first study to with LSMs and (2) following a play interaction with their explore the relationship between OT and different human fi infants, plasma OT levels showed a trend for a signi cant attachment relationships, Feldman et al. (2011) recruited 71 fi decrease in HSMs, whereas no signi cant change was seen in mothers and 41 fathers and their 4–6-month-old infants. Plasma, salivary, and urinary OT levels of parents were assessed before and after play interaction with infants. Higher plasma and salivary OT levels were found amongst parents who showed ‘high’ affect synchrony towards infants as compared with ‘low’ synchrony parents, supporting pre- vious findings by the same group that linked OT to more affectionate parenting behaviour (Feldman et al., 2007; Gordon et al., 2010). However, our findings are consistent with studies that implicate OT in the regulation of interpersonal stress in women (Taylor et al., 2006, 2010) including mothers (Feldman et al., 2011). In the latter study, among mothers only, post-interaction urinary OT levels (which did not correlate with plasma or salivary OT) were positively asso- ciated with anxiety in romantic attachments (i.e. relation- ships with a partner), with self-reported parenting stress, and with interactive stress (i.e. proportion of time when infant shows negative reactivity whilst mother tries to re-engage her/him during interactive play). Urinary OT levels were also reported to show a negative trend with own parenting care. Their findings suggest a role for OT in stress regulation in mothers which is also similar to that reported by other studies in women (Taylor et al., 2006, 2010; Turner et al., 2002). Riem et al. (2011) conducted the first randomised controlled trial to examine the effect of intranasal OT on brain responses (using fMRI) among women (non-mother) (21 women received OT and 21 received placebo). They reported that reduced activation in regions of aversion (amygdala) was seen in response to infant cry in women who received OT compared to those who received placebo. This supports the role of OT in social stress alleviation. In the current study, OT was assessed in mothers who were Fig. 2 – The relationship between post-interaction plasma specifically chosen because they occupied opposite ends of the oxytocin (OTpost) and own maternal overprotection among: distribution of maternal sensitivity; therefore, these mothers (a) low sensitivity (N 14) and (b) high sensitivity mothers may be more likely to have distinct affective and behavioural ¼ (N 15). caregiving attributes (Thompson, 1997). By classifying mothers ¼

Table 3 – Correlations between plasma OT and self-reported own parenting experience in mothers grouped by level of maternal sensitivity.

Own parenting experience domain HSMs (N 15) LSMs (N 14) ¼ ¼

OTpre OTpost OTpre OTpost

Maternal carea 0.38 0.21 0.21 0.07 À À À À Maternal overprotection 0.28 0.07 0.62n 0.63n Paternal carea 0.34 0.34 0.09 0.18 À À À Paternal overprotection 0.12 0.06 0.16 0.03 À À n po0.05. a Transformed variable. brain research 1580 (2014) 143–150 147

according to sensitivity behaviours, we may have ‘tapped’ into positive experience (viewing a comedy film), whereas no stress or anxiety coping strategies, at least in part modulated change was found after women viewed negative emotions by OT (see Numan and Woodside, 2010). As far as we are (viewing a sad film) (Turner et al., 2002). Recently, Strathearn aware, this is the first study to report such results in healthy et al. (2012) reported a drop in maternal plasma OT when mothers who have been carefully selected for high and low mothers showed higher levels of ‘effortful control’ during maternal sensitivity from within a healthy community sample. interaction with own infants, which is a prerequisite of OT appears to be an indicator of social affiliation (Feldman sensitive mothering. In another recent study, urinary OT in et al., 2007, 2010a, b), but it might also be a ‘signal’ for the mothers (n 26) was significantly higher following interaction ¼ need to affiliate with others (Tabak et al., 2011; Taylor et al., with an unfamiliar child (2.5–4.5 years) as compared with 2006, 2010). Animal literature suggests that OT has an ‘open- own child (Bick and Dozier, 2010). The authors concluded that ness’ to early social experience, with higher OT receptors interaction with an unfamiliar child might constitute a more density found in relation to enriched, early perceived parent- stressful situation that results in an increase in OT in order to ing environment (Champagne, 2008). The animal literature modulate this stress. (Numan and Woodside, 2010) also elucidates how OT serves a It is possible that HSMs perceive their infant signals as a dual role in maternal behaviour: It increases maternal moti- positive event (Turner et al., 2002), give appropriate attention vation and it also decreases stress and anxiety. The latter and focus to these signals (Strathearn et al., 2012), and that their effect may aid the mother in coping with difficult circum- plasma OT levels accordingly reflects this and falls during play stances related to infant care. With respect to the human with infant. By contrast, LSMs may not perceive interaction literature, different methods, which include postpartum with their infant as a positive event; they do not give proper stage of the mother–infant dyad and the ways in which attention and focus to their infant signals, and accordingly, mothers are classified, may differentially ‘tap into’ one or their OT levels remain relatively elevated. This is consistent other of these separable aspects of OT involvement in with Feldman et al. (2010a) who reported no difference in maternal care giving behaviour. plasma OT levels among mothers rated as low in affectionate contact following interaction with their infants. 3.1. Elevated plasma OT in LSMs 3.3. Strength and limitations Elevated plasma OT levels were previously reported in relation to social relationship difficulties. In a study assessing stress in As far as we are aware, this is the first study to chart differences partnerships among 85 adults in stable relationships (62% in plasma OT responses between mothers selected to represent women and 38% men), levels of plasma OT were significantly higher and lower extremes of maternal sensitivity in a popula- positively correlated with relationship distress in women (while tion of healthy new mothers. Its strengths include sampling plasma vasopressin correlated with relationship distress in healthy women from a community population and use of men) (Taylor et al., 2010). In a further study by the same group, rigorous methodology for ascertaining maternal sensitivity among 73 post menopausal women, plasma OT was negatively scores. Both the final (N 30) and the original sample (N 80) ¼ ¼ correlated with relationship with their own mother, or partner, provided for well-matched groups on key characteristics. and also marginally significantly correlated with relationship to There remain some important limitations. First, while the best friend (Taylor et al., 2006). Similarly, Marazziti et al. (2006) sample from which we derived the high and low sensitivity reported a positive correlation between plasma OT levels and groups was large enough to show a normal distribution of romantic relationship anxiety among 45 young subjects (33 sensitivity and variability among HSMs and LSMs (N 80), it ¼ women and 12 men). These findings suggest that plasma OT was still relatively small. Second, inferences about centrally increases as a ‘signal to affiliate with others as the pair-bond functioning OT from plasma measurement must remain relationship is threatened’.OTmight,therefore,actasa limited (Modahl et al., 1998) even if many previous studies biomarker for a distressed pair-bond relationship (Taylor do show modulation of peripheral plasma OT in relation to et al., 2010). In the current study, plasma OT among LSMs social affiliation (e.g. Gordon et al., 2008) or parenting brain was positively correlated with negative recall of own maternal responses (Strathearn et al., 2009). Third, the association we parenting experience, in particular, higher maternal overprotec- found between plasma OT and own maternal overprotection tion which is a marker of difficulties in the relationship with among HSMs does not imply a causal relationship, and there their own mothers. We did not assess the quality of the may be other, unidentified factors that could have accounted ‘relationship with partner’ and so we cannot exclude a dis- for this relationship. Fourth, own parenting experience was tressed pair-bond relationship as another explanation for the assessed by a self-report measure, although PBI has shown elevated plasma OT among the LSMs. good psychometric properties and convergent validity over 20 years (Wilhelm et al., 2005). Finally, we were unable to 3.2. Reductions in plasma OT in HSMs following infant examine for potential mediation effect that plasma OT might play play between perceived parenting experiences and maternal sensitivity towards own infant. This is because selection of Our findings suggest that plasma OT levels are reduced in the sample was based on maternal sensitivity which is the higher sensitivity mothers after playing with their infants but outcome variable for mediation analyses. that low sensitivity mothers showed no change in plasma Previous evidence suggested that in relation to plasma OT OT. A previous study has also demonstrated a reduction in and stress, levels were elevated only in the context of stimuli plasma OT in women (n 32) after a laboratory-induced which are social stressors i.e. social relationship difficulties. ¼ 148 brain research 1580 (2014) 143–150

However, we report evidence that a similar response occurs caregiver–infant interaction (Blazey et al., 2008; Murray et al., in new mothers with relatively reduced caregiving sensitivity. 1996). MACI has been validated for use in the 6–15 month-age Although in the current study we did not find a correlation range (Wan et al., 2012, 2013). The MACI-rated maternal between OT levels and depression scores, but this is not sensitivity has demonstrated reliability and moderately high necessarily inconsistent with our interpretation of the main stability over 6 months (r 0.48; Wan et al., 2013). ¼ findings since our sampling strategy meant that we excluded The rating of ‘sensitive responsiveness’ varies from (1) depressed mothers in the first phase and included only ‘minimally responsive/sensitive’ to (7) ‘very responsive/sen- extremes in maternal sensitivity in the second phase. sitive’. In the current study, high inter-rater agreement was Future studies replicating the present design might help- demonstrated on maternal sensitivity (intraclass correlation: fully combine measurements of plasma OT with measures r 0.70; po0.001, absolute agreement) based on independent ¼ assessing other aspects of the stress response axis (e.g. blind ratings of 30% of interaction clips in the complete cortisol). This might enable a more thorough evaluation of sample. Disagreements were resolved by both raters review- the role of OT in stress regulation (Neumann et al., 2000; ing the clips to reach consensus. Quirin et al., 2011). Plasma OT might usefully be employed as a biomarker in future studies exploring mechanisms of low 4.2.2. The Edinburgh postnatal depression scale (EPDS) parenting sensitivity and the effectiveness of interventions (Cox et al., 1987) designed to improve parenting quality. This 10-item self-report instrument is widely used to screen for depression in the postpartum and antenatal periods. Items are rated on a 4-point Likert scale and a cut-off score 4. Experimental procedures of 12 was used for screening positive. Using a cutoff point of 12, the EPDS shows sensitivity from 68% to 95% and a 4.1. Participants specificity of 78% to 96%, when compared to the diagnosis of major depression through psychiatric interview (Cox et al., Women were recruited from 6 community antenatal clinics 1987). across the northwest region of England (Greater Manchester), as part of a larger longitudinal study examining natural 4.2.3. The Hospital anxiety and depression rating scale variation in maternal sensitivity. Initially, 105 women, who (HADS) (Zigmond and Snaith, 1983) were ethnically white British with no psychiatric illness, and This 14-item questionnaire is a self-rating instrument to scored below the threshold on depression screening (see screen for anxiety and depression (7 items each). Items are below), were recruited during their last trimester of preg- rated on a 4-point Likert scale. Severity ratings correlate nancy (mean 33.90 weeks; SD 3.19). Following child birth, highly with clinical psychiatric assessments (r 0.70 depres- ¼ ¼ ¼ 80 women were then followed up and underwent evaluation sion and r 0.74 anxiety) (Zigmond and Snaith, 1983). The ¼ of maternal sensitivity using videoed mother–infant interac- measure has high internal consistency and high test-retest tion play at 4–6 months postpartum (mean 19.38 weeks; ¼ reliability (Crawford et al., 2001). In this study a cut-off score SD 2.47). Thirty mothers, representing extremes in the ¼ of 12 was used for screening positive. distribution of maternal sensitivity, were selected from this sample of 80. At 35 (mean) weeks postpartum (SD 3.26), we ¼ 4.2.4. Parental bonding instrument (PBI) (Parker et al., 1979) measured plasma OT levels before and after a mother–infant This self-report measure examines an adult's retrospective interaction. The majority of mothers (n 22, 73.3%) were on ¼ report of parents' caring behaviours (25 items for each parent) maternity leave. The study protocol was approved by the during the first 16 years of life, consisting of ‘care’ (12 items) North West Research Ethics Committee: Ref: 10/H1013/69 and and ‘overprotection’ (13 items). Items are rated on a 4-point written consent was obtained from all women. Likert scale. The PBI possessed good internal consistency (Parker et al., 1979). Examination of test-retest reliability over 4.2. Measures 20 years supports the construct and convergent validity of the measure over an extended period of time which was found to 4.2.1. Manchester assessment of caregiver–infant interaction be independent of mood effects (Wilhelm et al., 2005). (MACI) (Wan et al., 2012, 2013) This observational measure of caregiver–infant interaction 4.3. Procedure evaluates global features of interaction from brief unstructured play in 7 (2 caregiver, 3 infant, 2 dyadic) scales. The current 4.3.1. Time 1 and 2: pregnancy and mother–infant interaction study focused on the ‘caregiver sensitive responsiveness’ scale At the third trimester of pregnancy (Time 1), women com- (henceforth ‘maternal sensitivity’), defined as the “the extent to pleted PBI, EPDS and HADS. Mothers were excluded if they which the infant's moment-to-moment behaviour and devel- scored both EPDSZ12 and HADS-depression411 at any time opmental needs are responded to and supported by the point in the study. At 4–6 month postpartum (Time 2), caregiver, appropriately and contingently”.TheMACIwas mothers were visited at home and asked to play with their developed for research purposes to provide relatively brief infant on a floor mat as they normally do, with or without rating scales suitable for a wide age range in infancy, and toys (as supplied), as they wished. The interactions were which would provide measures of variance in the normal videotaped for 6 min and later rated, using the MACI caregiver population and be sensitive in at-risk samples. Its scales were sensitivity scale, independently by two trained researchers modified and refined from existing validated global scales of blind to study information. brain research 1580 (2014) 143–150 149

4.3.2. Time 3: oxytocin measurement Barrett, J., Fleming, A., 2011. Annual research review: all mothers Thirty mothers (15 HSMs and 15 LSMs) and their infants were are not created equal: neural and psychobiological invited to a clinical research facility for plasma OT measure- perspectives on mothering and the importance of individual ment either during mothers' menstruation or on contracep- differences. J. Child Psychol. Psychiatry 52, 368–397. Bick, J., Dozier, M., 2010. Mothers’ and children’s concentrations of tive pill free days. Blood samples were taken at the same time oxytocin following close, physical interactions with biological of day at 12:00–14:00 hours, an hour after the last nursing and non-biological children. Dev. Psychobiol. 52, 100–107. feed. Mothers were asked to refrain from caffeine and smok- Blazey, L., Leadbitter, K., Holt, C., Green, J., 2008. Attachment ing for at least 2 h beforehand. Three 5 ml samples of blood Behaviours and Parent–Child Interaction in Pre-school were taken from antecubital veins through an intravenous Autism, London, UK, International Meeting for Autism cannula. The first sample (OT1) was taken 10 min after Research, Poster. mother–infant separation, followed by reunion and a 10- Champagne, F., Diorio, J., Sharma, S., Meaney, M.J., 2001. Naturally minute mother–infant play interaction (as described earlier). occurring variations in maternal behavior in the rat are associated with differences in estrogen-inducible central The second and third samples were taken immediately post- oxytocin receptors. Proc. Natl. Acad. Sci. U. S. A. 98, interaction (OT2) and 5 min later (OT3). 12736–12741. Champagne, F. A, 2008. Epigenetic mechanisms and the 4.3.3. Oxytocin processing and assays transgenerational effects of maternal care. Front. All samples were processed as follows: samples were drawn Neuroendocrinol. 29, 386–397. into chilled vacutainer tubes containing lithium heparin Cox, J.L., Holden, J.M., Sagovsky, R., 1987. Detection of postnatal injected with 200 ml of Trasylol (aprotinin) 500,000 KIU/ml depression: development of the 10-item Edinburgh Postnatal Depression Scale. Br. J. Psychiatry 150, 782–786. blood. OT samples were kept ice-chilled until processed Crawford, J.R., Henry, J.D., Crombie, C., Taylor, E.P., 2001. within 10 min. Samples were then centrifuged at 4 1C at Normative data for the HADS from a large non-clinical 3500 rpm for 15 min. Five hundered microliters of super- sample. Br. J. Clin. Psychol. 40, 429–434. natants were transferred to 2 micro tubes (aliquot 1 and 2) Feldman, R., Weller, A., Zagoory-Sharon, O., Levine, A., 2007. and stored at 80 1C until transferred on dry ice to the Evidence for a neuroendocrinological foundation of human À University lab for analysis. OT was analysed using: ab13305- affiliation: plasma oxytocin levels across pregnancy and the Oxytocin ELISA (for protocol please refer to kithttp://www. postpartum period predict mother-infant bonding. Psychol. abcam.com/oxytocin-elisa-kit-ab133050.html). Determination Sci. 18, 965–970. Feldman, R., Gordon, I., Schneiderman, I., Weisman, O., of OT was performed using the Max Binding Determination Zagoory-Sharon, O., 2010a. Natural variations in maternal and Competitive Assay protocol on Gen 5 software using Biotek paternal care are associated with systematic changes in Plate reader. Oxytocin analyses were performed by a labora- oxytocin following parent-infant contact. tory scientist who was blind to all study information. Psychoneuroendocrinology 35, 1133–1141. Feldman, R., Gordon, I., Zagoory-Sharon, O., 2011. Maternal and 4.4. Statistical analyses paternal plasma, salivary, and urinary oxytocin and parent- infant synchrony: considering stress and affiliation components of human bonding. Dev. Sci. 14, 752–761. Demographic differences between the two groups of mothers Feldman, R., Zagoory-Sharon, O., Weisman, O., Schneiderman, I., were assessed through independent sample t-test or Gordon, I., Maoz, R., Shalev, I., Ebstein, R.P., 2012. Sensitive Chi-square test. Pearson correlations were used to examine parenting is associated with plasma oxytocin and correlations between plasma OT levels at the three assess- polymorphisms in the OXTR and CD38 genes. Biol. Psychiatry ment points as well as correlations with own parental 72, 175–181. experience. A repeated-measures analysis of variance Feldman, R., Gordon, I., Zagoory-Sharon, O., 2010b. The cross- (ANOVA) was employed to test for a significant change in generation transmission of oxytocin in humans. Hormon. plasma OT levels over time between groups (HSMs and LSMs). Behav. 58, 669–676. Francis, D.D., Champagne, F.C., Meaney, M.J., 2000. Variations in Analyses were performed using SPSS (version 19). maternal behaviour are associated with differences in oxytocin receptor levels in the rat. J. Neuroendocrinol. 12, 1145–1148. Acknowledgments Fries, A.B. W., Ziegler, T.E., Kurian, J.R., Jacoris, S., Pollak, S.D., 2005. Early experience in humans is associated with changes This study was supported by the Wellcome Trust Clinical in neuropeptides critical for regulating social behavior. Proc. Research Facility (Manchester), and the Centre for Women’s Natl. Acad. Sci. U. S. A. 102, 17237–17240. Mental Health (University of Manchester, UK). We are very Gordon, I., Zagoory-sharon, O., Schneiderman, I., Leckman, J.F., grateful to all the mothers and infants who have contributed Weller, A., Feldman, R., 2008. Oxytocin and cortisol in romantically unattached young adults: associations with to this study. bonding and psychological distress. Psychophysiology 45, 349–352. references Gordon, I., Zagoory-Sharon, O., Leckman, J.F., Feldman, R., 2010. Oxytocin and the development of parenting in humans. Bilol. Psychiatry 68, 377–382. Atzil, S., Hendler, T., Feldman, R., 2011. Specifying the Guzman, Y.F., Tronson, N.C., Jovasevic, V., Sato, K., Guedea, A.L., neurobiological basis of human attachment: brain, hormones, Mizukami, H., Nishimori, K., Radulovic, J., 2013. Fear- and behavior in synchronous and intrusive mothers. enhancing effects of septal oxytocin receptors. Nat. Neurosci. Neuropsychopharmacology 36, 1–13. 16, 1185–1187. 150 brain research 1580 (2014) 143–150

Lee, H.J., Macbeth, A.H., Pagani, J.H., Young, W.S., 2009. Oxytocin: Strathearn, L., Fonagy, P., Amico, J., Montague, P. R, 2009. Adult the great facilitator of life. Prog. Neurobiol. 88, 127–151. attachment predicts maternal brain and oxytocin response to Levine, A., Zagoory-Sharon, O., Feldman, R., Weller, A., 2007. infant cues. Neuropsychopharmacology 34, 2655–2666. Oxytocin during pregnancy and early postpartum: individual Strathearn, L., Iyengar, U., Fonagy, P., Kim, S., 2012. Maternal oxytocin patterns and maternal-fetal attachment. Peptides 28, response during mother-infant interaction: associations with 1162–1169. adult temperament. Hormon. Behav. 61, 429–435. Marazziti, D., Dell’Osso, B., Baroni, S., Mungai, F., Catena, M., Swain, J.E., Tasgin, E., Mayes, L.C., Feldman, R., Constable, R.T., Rucci, P., Albanese, F., Giannaccini, G., Betti, L., Fabbrini, L., Leckman, J.F., 2008. Maternal brain response to own baby-cry Italiani, P., Del Debbio, P., Lucacchini, A., Dell’Osso, L., 2006. is affected by cesarean section delivery. J. Child. Psychol. A relationship between oxytocin and anxiety of romantic Psychiatry 49, 1042–1052. attachment. Clin. Pract. Epidemiol. Mental Health 28. Tabak, B.A., McCullough, M.E., Szeto, A., Mendez, A.J., Modahl, C., Green, L., Fein, D., Morris, M., Waterhouse, L., McCabe, P.M., 2011. Oxytocin indexes relational distress Feinstein, C., Levin, H., 1998. Plasma oxytocin levels in autistic children. Biol. Psychiatry 43, 270–277. following interpersonal harms in women. Murray, L., Fiori-Cowley, A., Hooper, R., Cooper, P., 1996. The Psychoneuroendocrinology 36, 115–122. impact of postnatal depression and associated adversity on Taylor, S.E., Gonzaga, G.C., Klein, L.C., Hu, P., Greendale, G.A., early mother-infant interactions and later infant outcome. Seeman, S.E., 2006. Relation of oxytocin to psychological Child. Dev. 67, 2512–2526. stress responses and hypothalamic-pituitary-adrenocortical Neumann, I., Torner, L., Wigger, A., 2000. Brain oxytocin: axis activity in older women. Psychosom. Med. 68, 238–245. differential inhibition of neuroendocrine stress responses and Taylor, S.E., Saphire-Bernstein, S., Seeman, T.E., 2010. Are plasma anxiety-related behaviour in virgin, pregnant and lactating oxytocin in women and plasma vasopressin in men rats. Neuroscience 95, 567–575. biomarkers of distressed pair bond relationships?. Psychol. Numan, M., Stolzenberg, D.S., 2009. Medial preoptic area Sci. 21, 3–7. interactions with dopamine neural systems in the control of Thompson, R.A., 1997. Sensitivity and security: new questions to the onset and maintenance of maternal behavior in rats. ponder. Child. Dev. 68, 595–597. Front. Neuroendocrinol. 30, 46–64. Turner, R.A., Altemus, M., Yip, D.N., Kupferman, E., Fletcher, D., Numan, M., Woodside, B., 2010. Maternity: neural mechanisms, Bostrom, A., Lyons, D.M., Amico, J.A., 2002. Effects of emotion motivational processes, and physiological adaptations. Behav. on oxytocin, , and ACTH in women. Stress 5, 269–276. Neurosci. 124, 715–741. Wan, M.W., Green, J., Elsabbagh, M., Johnson, M., Charman, T., Parker, G., Tupling, H., Brown, L.B., 1979. Parental bonding Plummer, F., 2012. Parent-infant interaction in infant siblings instrument. Br. J. Med. Psychol. 52, 1–10. at risk of autism. Res. Dev. Disabil. 33, 924–932. Quirin, M., Kuhl, J., Dusing, R., 2011. Oxytocin buffers cortisol Wan, M.W., Green, J., Elsabbagh, M., Johnson, M., Charman, T., responses to stress in individuals with impaired emotion Plummer, F., 2013. The BASIS Team, Quality of interaction regulation abilities. Psychoneuroendocrinology 36, 898–904. Riem, M.M., Bakermans-Kranenburg, M.J., Pieper, S., Tops, M., between at-risk infants and caregiver at 12-15months is Boksem, M.A., Vermeiren, R.R., van Ijzendoorn, M.H., associated with 3-year autism outcome. J. Child. Psychol. Rombouts, S.A., 2011. Oxytocin modulates amygdala, insula, Psychiatry 54, 763–771. and inferior frontal gyrus responses to infant crying: a Wilhelm, K., Niven, H., Parker, G., Hadzi-Pavlovic, D., 2005. The randomized controlled trial. Biol. Psychiatry 70, 291–297. stability of the parental bonding instrument over a 20-year Ross, H.E., Young, L.J., 2009. Oxytocin and the neural mechanisms period. Psychol. Med. 35, 387–393. regulating social cognition and affiliative behavior. Front. Zigmond, A.S., Snaith, R.P., 1983. The hospital anxiety and Neuroendocrinol. 30, 534–547. depression scale. Acta Psychiatr. Scand. 67, 361–370. brain research 1580 (2014) 69–77

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Intranasal oxytocin effects on social cognition: A critique

Simon L. Evansa,1, Olga Dal Monteb,c,1, Pamela Nobleb, Bruno B. Averbeckb,n aSchool of Psychology, University of Sussex, Brighton, East Sussex BN1 9QG, United Kingdom bLaboratory of Neuropsychology, National Institute of Mental Health, National Institutes of Health, Bethesda, MD, USA cDepartment of Neuropsychology, University of Turin, Turin, Italy article info abstract

Article history: The last decade has seen a large number of published findings supporting the hypothesis Accepted 7 November 2013 that intranasally delivered oxytocin (OT) can enhance the processing of social stimuli and Available online 14 November 2013 regulate social emotion-related behaviors such as trust, memory, fidelity, and anxiety. The Keywords: use of nasal spray for administering OT in behavioral research has become a standard Oxytocin method, but many questions still exist regarding its action. OT is a peptide that cannot – Intranasal administration cross the blood brain barrier, and it has yet to be shown that it does indeed reach the brain CSF when delivered intranasally. Given the evidence, it seems highly likely that OT does affect Social stimuli behavior when delivered as a nasal spray. These effects may be driven by at least three Pro-social neuropeptide possible mechanisms. First, the intranasally delivered OT may diffuse directly into the CNS Anxiety where it directly engages OT receptors. Second, the intranasally delivered OT may trigger Social cognition increased central release via an indirect peripheral mechanism. And third, the indirect peripheral effects may directly lead to behavioral effects via some mechanism other than increased central release. Although intranasally delivered OT likely affects behavior, there are conflicting reports as to the exact nature of those behavioral changes: some studies suggest that OT effects are not always “pro-social” and others suggest effects on social behaviors are due to a more general anxiolytic effect. In this critique, we draw from work in healthy human populations and the animal literature to review the mechanistic aspects of intranasal OT delivery, and to discuss intranasal OT effects on social cognition and behavior. We conclude that future work should control carefully for anxiolytic and gender effects, which could underlie inconsistencies in the existing literature. This article is part of a Special Issue entitled Oxytocin and Social Behav. Published by Elsevier B.V.

1. Introduction neurotransmitter, and OT-like peptides can be found in nearly all vertebrate species (Gimple and Farenholtz, 2001). Peripheral Oxytocin (OT) is a peptide that has numerous functions in functions are wide in range. OT has a well-established role in the body, both peripherally as a hormone and centrally as a reproductive function (Corona et al., 2012; Courtois et al., 2013)

nCorresponding author. Fax: 1 301 402 0046. þ E-mail addresses: [email protected], [email protected] (B.B. Averbeck). 1S.L.E. and O.D.M. contributed equally and are co-first authors.

0006-8993/$ - see front matter Published by Elsevier B.V. http://dx.doi.org/10.1016/j.brainres.2013.11.008 70 brain research 1580 (2014) 69–77

and in parturition and lactation in females (Carson et al., 2013; Distribution patterns of OT receptors across brain areas are Gimple and Farenholtz, 2001). Synthetic OT has been used to highly species dependent (Insel and Shapiro, 1992; Young et al., assist in childbirth for decades. In addition, OT receptors are 1996), and binding sites are up-regulated in specificareasin located in visceral organs such as kidneys and pancreas, as well response to peripheral (such as pregnancy) or environmental as in the heart, fat cells, and adrenal glands (Gimple and (such as social cooperation) cues (Viero et al., 2010). Many OT Farenholtz, 2001), and OT has been found to be involved in neuron axonal projections run close to the ventricles, which may the regulation of water balance, density, and appetite allow for release of OT into the ventricles for communication (Carson et al., 2013). across numerous OT-receptive brain areas via the cerebrospinal In contrast, it has been suggested that OT effects in the fluid (CSF) (Veening et al., 2010). It has been proposed that this central nervous system (CNS) might be more specific, with OT global communication process through CSF within the CNS is playing an important role in modulating social behaviors and what may allow for the necessary simultaneous changes in the the processing of social stimuli. Whether these behavioral numerous neural mechanisms involved in rapid behavioral changes are modulated by OT in system-specific ways or due adaptation to environmental stimuli (Veening et al., 2010). to more general effects are, however, unknown. The study of central effects of OT has been carried out in animal models and humans using different delivery methods: in animals 2.1. CSF versus plasma both central and peripheral administration has been used, while in humans studies investigating the effects of exogen- The relationship between peripheral (plasma) and central ous OT typically use intranasal spray for delivery, with few (CSF) OT levels is complex. Studies in humans are typically exceptions (Hollander et al., 2003). How or if the OT enters the restricted to peripheral OT assessments due to the risks brain using this method is, however, still unknown. The associated with invasive CSF collection procedures (Carter, purpose of this critique is twofold. We firstly discuss the 2007; Challinor et al., 1994; Domes et al., 2010; Modahl et al., potential mechanisms by which OT could enter the brain, 1998; Parker et al., 2010). Studies that use measures in plasma and weigh the evidence from work in animals. Implications to track changes in OT levels after nasal administration in for human studies using intranasal OT are discussed. We humans have found significant increases in OT levels from then provide an overview of intranasal OT effects on social baseline at 30 min (Gossen et al., 2012), 45 min (Domes et al., cognition in healthy humans, and explore whether OT 2010), and over the course of 1 h (Burri et al., 2008). Further- is genuinely a neuropeptide with specifically “pro-social” more, a number of human studies have reported correlations effects. We incorporate findings published since other recent between peripheral levels of endogenous OT and behavior. reviews on this topic (Bartz et al., 2011; Guastella et al., 2013; High levels of plasma OT have been associated with trust and MacDonald et al., 2011), identify potential confounds that trustworthiness (Zak et al., 2005, 2007), positive physical could underlie current inconsistencies in the literature, and contact with a partner (Grewen et al., 2005), and lower levels provide suggestions as to how these could be resolved. In of anxiety in patients with depression (Scantamburlo et al., tying together both the mechanistic and behavioral aspects of 2007). By contrast, low peripheral levels of OT have been intranasal OT delivery, we provide a summary several issues found in patients with depression (Cyranowski et al., 2008), as a guide for future research. schizophrenia, (Goldman et al., 2008; Keri et al., 2009) and autism spectrum disorders (Green et al., 2001). However, there are numerous animal studies that show no correlation between plasma and CSF levels of OT in response to 2. Intranasal delivery: mechanisms avarietyofmanipulations,rangingfromhormoneadministra- tion to environmental cues (Amico et al., 1990; Rosenblum et al., The OT peptide is composed of nine amino acids and is 2002; Seckl and Lightman, 1987; Veening et al., 2010; Winslow produced in the paraventricular nucleus (PVN) and supraoptic et al., 2003). Winslow et al. (2003) reported significant differences nucleus (SON) of the hypothalamus in mammals. OT is in endogenous CSF OT levels between nursery and mother released peripherally primarily from the neurohypophysis reared rhesus monkeys over a period of development, but no by exocytosis (Carson et al., 2013; Viero et al., 2010). Since difference in plasma OT between the groups, and no correlation OT is a relatively large, hydrophilic molecule, blood–brain between CSF and plasma OT. Because OT cannot cross the penetration is too poor to cause any measurable effects on blood–brain barrier (BBB), central and peripheral OT systems central systems (McEwen, 2004), so peripheral OT likely may be independently regulated, thus peripheral and central re-enters the brain in negligible amounts. Instead, OT is released effects of OT are thought to be coordinated through its common directly in the CNS by OT neurons that project to numerous release as a result of collateral axons in the pituitary and brain regions from the PVN, separate from those that go to the nucleus accumbens (Ross and Young, 2009). It may be that pituitary (Ross and Young, 2009; Veening et al., 2010). correlations cannot be detected due to temporal differences in OT receptors are widely distributed through many brain areas responses between peripheral and central systems (Neumann in rat, including the spinal cord, brainstem, hypothalamus, et al., 2013). Bioavailability differs significantly in plasma versus amygdala, and nucleus accumbens (Ross and Young, 2009). CSF; OT is broken down within 2 min in plasma, while it lasts for While localization of OT receptors has yet to be definitively up to a half an hour in CSF due to a lack of hydrolyzing enzymes mapped in primates and humans (Toloczko et al., 1997), efforts (Jones and Robinson, 1982; Mens et al., 1983; Robinson et al., are being made to develop a radioligand that will bind with 1982; Robinson and Coombes, 1993; Veening et al., 2010; Viero high specificity to human OT receptors (Smith et al., 2012). et al., 2010). brain research 1580 (2014) 69–77 71

Importantly, whether peripheral OT levels are indicative of Corresponding human data does not exist for OT, as central OT levels remains an open question. It is interesting clinical studies are limited to assessing only plasma levels that both CSF and plasma OT levels can be correlated with after intranasal delivery (but see Born et al., 2002, for behavioral changes, but do not have a detectable correlation vasopressin assessment in CSF), though these studies con- with each other. More work needs to be done to further sistently find correlations with increased plasma OT levels investigate the nature of the relationship between peripheral and changes in behavior (Challinor et al., 1994; Domes et al., and central levels. 2010). OT behavioral studies often cite work showing that vasopressin, which is closely related to OT, reaches peak levels in CSF in 30-50 min when administered intranasally 2.2. Intranasal delivery (Born et al., 2002). Limited data has shown increased levels of central OT after nasal administration, in two rhesus maca- There is considerable uncertainty as to how intranasal OT might ques (Chang et al., 2012). Future studies with a larger group of exert behavioral effects. Animal studies suggest that intranasal subjects are planned to determine whether intranasal admin- delivery can bypass the BBB via extracellular pathways in the istration of OT leads to reliable elevation of OT in the CSF and epithelium, though more than one route is possible. Two plasma compared to placebo. potential pathways have been identified: a peripheral olfactory While there is some suggestion that intranasal administra- route connecting the nasal passages with the olfactory bulbs and tion of OT increases CSF OT levels, the route of action is still rostral brain regions (e.g. anterior olfactory nucleus and frontal unknown. It is not clear whether behavioral effects that follow cortex), and a peripheral trigeminal system connecting the nasal peripheral administration of OT are driven by increased passages with brainstem and spinal cord regions. Both routes central concentrations of OT due to the OT entering the CNS, provide rapid entry to the CNS (Thorne et al., 2004). Given that or to an indirect peripheral effect driving central production the behavioral effects of intranasal OT can appear shortly after and release of OT. A third possibility is that peripheral effects delivery, these routes are likely candidates for a transport of OT may drive indirect behavioral effects through unknown mechanism by which OT could enter the brain. Recent rodent mechanisms. Further studies will be needed to determine the work supports this suggestion (Neumann et al., 2013). After nasal route and distribution of OT following intranasal delivery. administration of OT in rats and mice, it has been shown using There are at least two possible experimental approaches to microdialysis that OT levels increase in the dorsal hippocampus track the intranasal route into the CNS. First, using a periph- and amygdala of rats and mice, with peak levels occurring eral OT antagonist that does not cross the BBB in conjunction 30–60 min after administration. Corresponding changes were with intranasal OT administration would eliminate any per- also observed in plasma. OT was found to be quite uniformly ipheral actions (blocking peripheral OT receptors) that might distributed within the brain extra cellular fluid (ECF) suggesting confound interpretation of central effects. Second, radiolabel- that effects were not due to locally released endogenous OT, ling or fluorophore-conjugated OT delivered in nonhuman particularly because there are no OT receptors in the dorsal primates would reveal the pattern of uptake and distribution, hippocampus and no local OT pathways terminate in that area and show whether it is the exogenous OT that is acting (Neumann et al., 2013). This is of note, since it has been directly in central locations, or whether it is triggering endo- speculated that very small volumes of the intranasally delivered genous production and release without uptake into central drug might be sufficient to trigger release of endogenous OT as areas. exogenous OT has been shown to have a positive-feedback effect Finally, it should be noted that individual differences in on endogenous release in a dose-dependent fashion (Falke, 1989; anatomy likely introduce variation in intranasal OT uptake. Moos et al., 1984). However, the lowest dose of OT applied in this Nasal anatomy influences airflow, which influences the study was about 400 pg/ml. Estimates of the CSF level of OT in degree to which OT accesses the epithelium following nasal rats and monkeys following intranasal delivery is about 50 pg/ml spray. Whether other constituents are included in the man- (Neumann et al., 2013; Chang et al., 2012). Thus, estimates of CSF ufacture of the spray (e.g. uptake enhancers) is important, levels achieved by intranasal delivery are lower than the levels and dosage has also varied considerably between studies. (400 pg/ml) that have been shown to lead to endogenous release The confounding effects of these factors could explain some and therefore it is not clear if the intranasally delivered OT can of the inconsistencies in findings that are discussed in the drive this mechanism. next section. To minimize these effects, a protocol, which Work with other peptides shows consistent findings. standardizes intranasal OT administration, has been pro- Neuropeptide S (NPS), another large, hydrophilic molecule, posed by Guastella et al. (2013). is detectable in the rat brain using fluorophore-conjugation techniques at 15 min after its intranasal delivery. It was also found that the neuronal populations targeted by intranasal 3. Intranasal oxytocin: effects on social administration were identical to those targeted by intracer- cognition ebroventricular injection of NPS into the right lateral ventricle (Ionescu et al., 2012), suggesting that either delivery method A possible “pro-social” role for OT was first suggested by work would lead to similar results. Also, intranasal administration showing that OT could induce maternal behaviors in virgin of radiolabeled 60-amino acid galanin-like peptide has been female rats (Pedersen and Prange, 1979). Subsequent work shown to cause transport of the peptide into the rat olfactory showed that pulsatile release of OT influences pair bonding in bulb, as well as other selected brain regions when combined the monogamous female prairie vole, in combination with with cyclodextrins for specific targeting (Nonaka et al., 2008). olfactory signals (Williams et al., 1994). These findings, along 72 brain research 1580 (2014) 69–77

with the availability of OT nasal sprays, inspired researchers evidence that OT can enhance amygdala activity for happy to search for corresponding effects in humans. The last expressions, possibly reflecting a shift of processing toward decade or so has seen a large number of published findings positive social stimuli (Gamer et al., 2010). Whether these demonstrating that OT can enhance the processing of social effects generalize to females is unclear; the issue of gender stimuli and regulate various social behaviors. effects is discussed further below.

3.1. Facial expressions 3.2. Trust

OT has been shown to influence the recognition of emotional One of the most highly publicized effects of OT is its role in facial expressions in static (Di Simplicio et al., 2009; Guastella trust. Kosfeld et al. (2005) used a trust game involving real et al., 2010; Marsh et al., 2010) and dynamic (Fischer-Shofty monetary stakes, where participants were asked to play the et al., 2010; Lischke et al., 2012a) images of faces. In addition, role of either investor or trustee. Investors who had been studies have shown that OT affects probabilistic learning treated with OT made significantly more investments com- when participants have to learn to associate reward with pared to investors treated with placebo. Importantly, no happy, sad or angry faces: OT appears to counteract the difference was seen in an identical ‘risk’ task where partici- natural aversion to selecting angry face stimuli while having pants were playing against a computer rather than another no effects on other expressions (Evans et al., 2010). Other human. OT has also been shown to affect trust adaptation: work has shown that OT can introduce a general positive bias OT-treated individuals have been shown to persist in trusting in face judgment. OT administration increases ratings of behaviors even after repeated betrayal, and this has been trustworthiness and attractiveness of male and female tar- linked to a modulation of amygdala activity (Baumgartner gets in raters of both sexes (Theodoridou et al., 2009), and et al., 2008). Conscious and implicit distrust of faces is some researchers have found that OT enhances emotion associated with amygdala activation (Winston et al., 2002) recognition only for positive expressions (Marsh et al., 2010) while amygdala lesions increase trust (Adolphs et al., 1998). and slows recognition of fearful ones (Di Simplicio et al., Thus amygdala deactivation could explain why OT enhances 2009), although conflicting results have been reported, with stimulus processing and promotes trust behavior. some studies reporting effects for fearful expressions only However, other findings contradict the notion that OT reliably (Fischer-Shofty et al., 2010). Work by Lischke et al. (2012b) promotes “pro-social” behavior (Bosch et al., 2005; Shamay- found that OT reduced the intensity at which all emotions Tsoory et al., 2009). For example, findings by Shamay-Tsoory could be recognized, with some evidence that this effect et al. (2009) seem to suggest that OT effects are not always favors angry and fearful expressions. Several of these results positive in a trust context. Using a game of chance with financial are inconsistent with respect to which emotions are differ- stakes, it was found that OT increased self-report ratings of envy entially processed following OT administration. This may be when another (fake) participant won more money, and increased due to the small number of participants often used, and ratings of gloating when the other participant lost more money. differences in the experimental approaches. These results are presented as evidence for OT effects on a wide Studies of covert attention have shown that OT creates an range of (not necessarily positive) social emotion-related beha- attentional bias towards happy over angry faces, indicating viors. However, an alternative view is that OT reduces social that OT might shift early attentional processes towards inhibition, thus making participants more likely to admit to positive social stimuli (Domes et al., 2013). Eye-tracking socially unacceptable emotions such as gloating. studies with male participants have shown that OT increases gaze time spent exploring the eye region compared with 3.3. Memory other parts of a face (Andari et al., 2010; Gamer et al., 2010; Guastella et al., 2008a), suggesting that improvements in OT also seems to promote recognition memory for social facial emotion recognition might be due to an enhanced stimuli. OT administered before encoding of face stimuli processing of the information-rich eye region. Three other improves the accuracy of familiarity made a day later, by studies, however, have not replicated this finding in female lowering the false alarm rate and thereby improving the participants (Domes et al., 2010; Lischke et al., 2012a, 2012b). signal-to-noise ratio for discriminating new faces from old In males, OT enhances the ability to infer the mental states (Rimmele et al., 2009). Other studies have shown similar of another person when only the eye regions are presented as a results, although Savaskan et al. (2008) found that OT cue (Domes et al., 2007; Guastella et al., 2010). OT has been improved recognition memory for neutral and angry, but shown to decrease amygdala activation to fearful, angry and not happy faces, whereas Guastella et al. (2008b) found that happy faces even when presented implicitly (Domes et al., the effect was only present for happy faces. Despite this 2007), suggesting that an OT-mediated reduction in amygdala inconsistency, Rimmele et al. showed that OT does not activation might encourage theperceptionofsocialcuesby improve recognition for non-social stimuli. It has been argued making participants feel more at ease while viewing faces, since that this result parallels similar findings in mice, since perception of any face stimulus activates the amygdala, parti- OT-knockout mice seem to be impaired at detecting whether cularly if its unfamiliar (Haxby et al., 2000)orinvolvesdirecteye an intruder is novel or familiar, while non-social memory and contact (Kawashima et al., 1999). After intranasal OT, decreased olfactory function appear intact (Ferguson et al., 2002), and activity in amygdala sub-regions associated with face proces- OT injected into the amygdala can reinstate normal perfor- sing seems to be associated with increases in sub-regions mance (Ferguson et al., 2001). However, as noted by Insel and associated with gaze, supporting this suggestion; there is also Fernald (2004), mouse social recognition paradigms differ brain research 1580 (2014) 69–77 73

from person recognition in humans in that pheromonal to social and non-social threat, an effect which could be recognition is a major factor. mediated by estrogen (Domes et al., 2010; Lischke et al., 2012b). More work is needed to investigate these gender 3.4. Relationship status differences; this is particularly relevant when considering OT in a therapeutic context. There is intriguing evidence that OT might promote fidelity within monogamous human relationships and thus carry out a 3.6. Effects in clinical populations function similar to that observed in prairie voles. OT adminis- tration causes men in a monogamous relationship, but not Despite considerable interest in the “pro-social” effects of OT single ones, to distance themselves from an attractive female as a potential treatment option for schizophrenia and autism, experimenter, and to approach pictures of attractive women it is important to note that OT does not always produce more slowly (Scheele et al., 2012). The early stages of romantic positive social effects in clinical populations. It has been attachment have been linked to enhanced OT levels in plasma, shown that the “typical” responses to acute intranasal the levels of which can predict whether the relationship is administration can be altered by individual variation in sustained 6 months later (Schneiderman et al., 2012). Physical psychological profile or endocrine systems, and dose levels. intimacy and greater spousal support in a relationship has been OT can actually hinder trust and cooperation in participants linked to higher plasma OT levels (Grewen et al., 2005; Light et al., with Borderline Personality Disorder (Bartz et al., 2010). 2005), and it has been shown that a dose of intranasal OT A dose-dependent result in identifying emotional expressions delivered prior to a couple conflict discussion can increase in schizophrenic patients was found by Goldman et al. (2011), positive communication and reduce conflict-related rises in showing that emotion recognition actually decreased from salivary cortisol levels (Ditzen et al., 2009). Thus higher basal baseline at a low dose of 10 IU, but then improved above OT levels, and intranasal OT, seem to promote attachment to a baseline at 20 IU. Bales et al. (2013) also found dose- partner and the resolution of conflicts, and OT causes males in a dependent effects on the development of pair-bonding in monogamous relationship to behave cautiously in the presence male voles; the direction of OT effects differed according to of another female, promoting monogamy. Interestingly, other whether it was delivered acutely or chronically, demonstrat- studies have shown that plasma OT also increases with relation- ing further that dosing regimen might be critical. The “pro- ship distress in women (Taylor et al., 2006, 2010), and this is social” effects outlined above might not generalize to a supported by work in prairie voles showing that long-term social chronic administration schedule. isolation causes elevated plasma OT in females (Grippo et al., 2007). Elevated OT in this context has been interpreted as a signal to affiliate with others because the pair–bond relationship 3.7. Specificity of OT effects is threatened (Taylor et al., 2010). This issue requires further exploration, as does the gender-specificity. It would also be A criticism of the work outlined above concerns the issue of interesting to investigate whether relationship status affects specificity. The animal literature argues convincingly for OT response to OT within other social settings, such as trust. effects to be specifically social in nature. Studies in humans have tended to work from the assumption that this is also the 3.5. Inconsistent gender effects case, to the exclusion of considering alternative explanations for their findings. Churchland and Winkielman (2012) argue As discussed briefly above, both animal and human studies that OT effects might be better characterized in terms of more have reported inconsistent responses to OT between male general mechanisms such as anxiety reduction, affiliative and female subjects. Animal studies have suggested that the motivation, or social saliency. Indeed, it would be unusual effects of OT might be modified by interactions with estrogen for a hormone and neurotransmitter such as OT to affect (McCarthy et al., 1996), and many human studies have opted processing and responses under very particular circum- to include male participants to simplify interpretation of stances, especially when such circumstances are those which results. Significant sex specific effects have been found for require the relatively abstract, ‘higher order’ operations of the vasopressin (AVP). In men, intranasally administered AVP social brain. Churchland and Winkielman instead suggest stimulates agonistic facial expressions and decreases percep- that the anxiolytic effects of OT can explain the majority of tion of trust in response to pictures of same-sex strangers. In findings. These anxiolytic effects have been well demon- women, administration of the same peptide results in affilia- strated in mice (Mantella et al., 2003; Ring et al., 2006) and tive facial expressions and increased perception of friendli- there is some evidence for similar effects in humans. Nursing ness (Thompson et al., 2006). Although limited data is mothers with higher OT plasma levels are more likely to available, studies point to powerful gender differences in OT describe positive mood states and reduced anxiety (Carter effects as well. A recent study has investigated the role of et al., 2001; Heinrichs et al., 2001) whereas abuse in childhood exogenous OT in reaction to social stress in both genders has been linked to lower OT concentrations in CSF and higher (Kubzansky et al., 2012). Males given OT reported less distress anxiety scores (Heim et al., 2009). Intranasal OT has been to stress exposure, while females reported more distress and shown to decrease anxiety in a simulated public speaking more anger. Moreover, neuroimaging work suggests diver- test (de Oliveira et al., 2012; Heinrichs et al., 2003). The gent effects on amygdala activation. In males, OT tends to findings described above, which show that OT acts to pro- elicit decreased amygdala activity in response to emotional mote affiliative social behaviors, could be due to a general faces (Domes et al., 2007); in females, OT enhances reactivity reduction in anxiety. However, studies that have attempted 74 brain research 1580 (2014) 69–77

to assess mood and anxiety changes using brief self-report It is interesting to note that OT effects on the processing of questionnaires after intranasal OT tend not to find any social stimuli seem to be more robust than effects on social changes (Guastella et al., 2008b). This is confirmed by a behaviors. In contrast to the work by Kosfeld et al. described systematic review of 38 randomised controlled trials of above, some studies have shown that OT can actually decrease intranasal OT, which concluded that OT produces no detect- trust behaviors if the other party is portrayed as untrustworthy able subjective changes in recipients (MacDonald et al., 2011). (Mikolajczak et al., 2010), is unknown (Declerck et al., 2010)orisa Nevertheless, the notion that anxiolytic effects must be member of a social out-group (De Dreu et al., 2010). It could be present to some degree is supported by the OT neuroimaging that OT reliably increases the salience of social stimuli, but “pro- data, the majority of which concerns the amygdala, a key site social” behaviors only emerge in the presence of context- in anxiety regulation. A study by Singer et al. (2008) investi- dependent anxiolytic effects, or be dependent on other factors. gated OT effects on empathy (using a paradigm where parti- Researchers should therefore be wary of assuming that OT cipants observe, and receive, painful stimulation of the hand), enhances all aspects of social function. This is demonstrated and “pro-social” behavior (using an economic exchange para- by Hurlemann et al. (2010),whofoundthatwhileOTincreased digm). The only observed effects were a reduction in amygdala self-reported emotional empathyintheMultifacetedEmpathy activation in response to painful stimulation of self; “pro- Test, OT did not improve a participant’saccuracyforinferring social” behavior and empathy were unaffected, as was activity mental states. This finding supports the notion that OT makes in brain regions implicated in empathy tasks of this type, such social stimuli more salient, presumably at the level of the as insula. amygdala, but this enhancement does not necessarily translate In contrast, there are some lines of evidence suggesting to improved theory of mind. OT effects on higher-order social that OT effects are indeed specifictosocialcontextsandnot function seem to be complex and more work is certainly needed anxiety-dependent. The use of non-social controls demon- to determine the circumstancesunderwhichintranasalOTis strates specificity of effects, although it could be argued that “pro-social”. some other variable (e.g. perceived complexity or importance) is not identical across conditions and thus responsible for the lack of effects (Kosfeld et al., 2005). It is hard to level such criticism at 4. Conclusion astudybyUnkelbach et al. (2008),whofoundthatOTspeeds detection of positive words specifically associated with sexu- In conclusion, there are still many questions regarding the ality, bonding, and social relationships while having no effects mechanisms by which intranasal delivery of OT enters on other positive and negative stimuli. Pupillometry studies the brain and the inconsistent behavioral effects reported in have shown greater cognitive resource allocation to social the literature. Future studies with larger subject groups should stimuli under OT (Prehn et al., 2013). Hurlemann et al. (2010) investigate whether intranasal administration of OT leads to used social (happy/angry faces) and non-social (colored dots) reliable elevation of OT in the CSF and plasma and if periph- reinforcements in an associative learning task to show that eral and central OT levels are correlated. Furthermore, more OT improves learning in the social condition only. OT also studies will be needed to determine the route and distribution improved self-reported emotional empathy intensity ratings for of OT following intranasal delivery. Despite the multitude of both positively and negatively valence pictures. studies investigating the effects of OT on human behavior and Perhaps the evolutionary importance of our social world social cognition, the inconsistent results leave open the debate justifies an argument for the effect of OT being socially specific. between the notion that OT reliably promotes “pro-social” It is possible that a more general mechanism could explain the behavior or has anxiolytic effects. Moreover, some studies findings, but since the literature is so heavily biased towards have pointed out powerful gender differences in OT effects, social processing thereisatpresentinsufficient data. However, but most studies were conducted in males only, discounting it is important to note that the anxiolytic effects of OT, which gender effects. It seems that the small numbers of male have been put forward as a possible explanatory mechanism participants typically employed, combined with differences (Churchland and Winkielman, 2012), are poorly specified in in methods, tasks, and the stimulus sets used, could underlie humans. Evidence of anxiolyticeffectshasbeeninferredfrom these inconsistent findings. Future studies should investigate studies measuring basal levels of plasma OT (and thus offers no OT effects across genders and be designed to determine the proof of a direct relationship, as well as being subject to the degree to which possible general anxiolytic effects contribute criticisms outlined above), and work showing an anxiolytic to changes in response in socially specific challenges. effect in a social context. Could it be that anxiolytic effects only emerge in a social setting, with anxiolysis actually resulting from a positive processing bias for social stimuli, generating an Funding enhanced of social approval? In line with this suggestion, OT has been shown to potentiate the anxiolytic effect of social This research was supported by the Intramural Research support during a stressful public speaking test (Heinrichs et al., Program of the National Institute of Mental Health, NIMH. 2003). It is clear that the anxiolytic effects of intranasal OT need to be better characterized, and all future studies should take references more care in excluding anxiolytic effects as a confound. The literature would also benefitfromawideruseofnon-social control conditions, to assert specificity of effects with more Adolphs, R., Tranel, D., Damasio, A.R., 1998. The human amygdala confidence. in social judgment. Nature 393, 470–474. brain research 1580 (2014) 69–77 75

Amico, J.A., Challinor, S.M., Cameron, J.L., 1990. Pattern of de Oliveira, D.C., Zuardi, A.W., Graeff, F.G., Queiroz, R.H., Crippa, oxytocin concentrations in the plasma and cerebrospinal fluid J.A., 2012. Anxiolytic-like effect of oxytocin in the simulated of lactating rhesus monkeys (Macaca mulatta): evidence for public speaking test. J. Psychopharmacol. 26, 497–504. functionally independent oxytocinergic pathways in primates. Declerck, C.H., Boone, C., Kiyonari, T., 2010. Oxytocin and J. Clin. Endocrinol. Metab. 71, 1531–1535. cooperation under conditions of uncertainty: the modulating Andari, E., Duhamel, J.R., Zalla, T., Herbrecht, E., Leboyer, M., role of incentives and social information. Horm. Behav. 57, Sirigu, A., 2010. Promoting social behavior with oxytocin in 368–374. high-functioning autism spectrum disorders. Proc. Natl. Acad. Di Simplicio, M., Massey-Chase, R., Cowen, P.J., Harmer, C.J., 2009. Sci. U. S. A. 107, 4389–4394. Oxytocin enhances processing of positive versus negative Bales, K.L., Perkeybile, A.M., Conley, O.G., Lee, M.H., Guoynes, emotional information in healthy male volunteers. C.D., Downing, G.M., Yun, C.R., Solomon, M., Jacob, S., J. Psychopharmacol. 23, 241–248. Mendoza, S.P., 2013. Chronic intranasal oxytocin causes long- Ditzen, B., Schaer, M., Gabriel, B., Bodenmann, G., Ehlert, U., term impairments in partner preference formation in male Heinrichs, M., 2009. Intranasal oxytocin increases positive prairie voles. Biol. Psychiatry 74, 180–188. communication and reduces cortisol levels during couple Bartz, J.A., Zaki, J., Bolger, N., Hollander, E., Ludwig, N.N., conflict. Biol. Psychiatry 65, 728–731. Kolevzon, A., Ochsner, K.N., 2010. Oxytocin selectively Domes, G., Heinrichs, M., Glascher, J., Buchel, C., Braus, D.F., improves empathic accuracy. Psychol. Sci. 21, 1426–1428. Herpertz, S.C., 2007. Oxytocin attenuates amygdala responses Bartz, J.A., Zaki, J., Bolger, N., Ochsner, K.N., 2011. Social effects of to emotional faces regardless of valence. Biol. Psychiatry 62, oxytocin in humans: context and person matter. Trends Cogn. 1187–1190. Sci. 15, 301–309. Domes, G., Lischke, A., Berger, C., Grossmann, A., Hauenstein, K., Baumgartner, T., Heinrichs, M., Vonlanthen, A., Fischbacher, U., Heinrichs, M., Herpertz, S.C., 2010. Effects of intranasal Fehr, E., 2008. Oxytocin shapes the neural circuitry of trust oxytocin on emotional face processing in women. and trust adaptation in humans. Neuron 58, 639–650. Psychoneuroendocrinology 35, 83–93. Born, J., Lange, T., Kern, W., McGregor, G.P., Bickel, U., Fehm, H.L., Domes, G., Sibold, M., Schulze, L., Lischke, A., Herpertz, S.C., 2002. Sniffing neuropeptides: a transnasal approach to the Heinrichs, M., 2013. Intranasal oxytocin increases covert human brain. Nat. Neurosci. 5, 514–516. attention to positive social cues. Psychol. Med. 43, 1747–1753. Bosch, O.J., Meddle, S.L., Beiderbeck, D.I., Douglas, A.J., Neumann, Evans, S., Shergill, S.S., Averbeck, B.B., 2010. Oxytocin decreases I.D., 2005. Brain oxytocin correlates with maternal aggression: aversion to angry faces in an associative learning task. link to anxiety. J. Neurosci. 25, 6807–6815. Neuropsychopharmacology 35, 2502–2509. Burri, A., Heinrichs, M., Schedlowski, M., Kruger, T.H., 2008. The Falke, N., 1989. Oxytocin stimulates oxytocin release from acute effects of intranasal oxytocin administration on isolated nerve terminals of rat neural lobes. Neuropeptides 14, endocrine and sexual function in males. 269–274. Psychoneuroendocrinology 33, 591–600. Ferguson, J.N., Aldag, J.M., Insel, T.R., Young, L.J., 2001. Oxytocin in Carson, D.S., Guastella, A.J., Taylor, E.R., McGregor, I.S., 2013. the medial amygdala is essential for social recognition in the A brief history of oxytocin and its role in modulating mouse. J. Neurosci. 21, 8278–8285. psychostimulant effects. J. Psychopharmacol. 27, 231–247. Ferguson, J.N., Young, L.J., Insel, T.R., 2002. The neuroendocrine Carter, C.S., Altemus, M., Chrousos, G.P., 2001. Neuroendocrine basis of social recognition. Front Neuroendocrinol. 23, and emotional changes in the post-partum period. Prog. Brain 200–224. Res. 133, 241–249. Fischer-Shofty, M., Shamay-Tsoory, S.G., Harari, H., Levkovitz, Y., Carter, C.S., 2007. Sex differences in oxytocin and vasopressin: 2010. The effect of intranasal administration of oxytocin on implications for autism spectrum disorders?. Behav. Brain fear recognition. Neuropsychologia 48, 179–184. Res. 176, 170–186. Gamer, M., Zurowski, B., Buchel, C., 2010. Different amygdala Challinor, S.M., Winters, S.J., Amico, J.A., 1994. Pattern of oxytocin subregions mediate valence-related and attentional effects of concentrations in the peripheral blood of healthy women and oxytocin in humans. Proc. Natl. Acad. Sci. U. S. A. 107, men: effect of the and short-term fasting. 9400–9405. Endocr. Res. 20, 117–125. Gimple, G., Farenholtz, F., 2001. The oxytocin receptor system: Chang, S.W., Barter, J.W., Ebitz, R.B., Watson, K.K., Platt, M.L., 2012. structure, function, and regulation. Physiol. Rev. 81, 629–683. Inhaled oxytocin amplifies both vicarious reinforcement and Goldman, M., Marlow-O’Connor, M., Torres, I., Carter, C.S., 2008. self reinforcement in rhesus macaques (Macaca mulatta). Diminished plasma oxytocin in schizophrenic patients with Proc. Natl. Acad. Sci. U. S. A. 109, 959–964. neuroendocrine dysfunction and emotional deficits. Churchland, P.S., Winkielman, P., 2012. Modulating social Schizophr. Res. 98, 247–255. behavior with oxytocin: how does it work? What does it Goldman, M.B., Gomes, A.M., Carter, C.S., Lee, R., 2011. Divergent mean?. Horm. Behav. 61, 392–399. effects of two different doses of intranasal oxytocin on facial Corona, G., Jannini, E.A., Vignozzi, L., Rastrelli, G., Maggi, M., 2012. affect discrimination in schizophrenic patients with and The hormonal control of ejaculation. Nat. Rev. Urol. 9, without polydipsia. Psychopharmacology (Berl). 216, 101–110. 508–519. Gossen, A., Hahn, A., Westphal, L., Prinz, S., Schultz, R.T., Courtois, F., Carrier, S., Charvier, K., Guertin, P.A., Journel, N.M., Grunder, G., Spreckelmeyer, K.N., 2012. Oxytocin plasma 2013. The control of male sexual responses. Curr. Pharm. Des. concentrations after single intranasal oxytocin 19, 4341–4356. administration-A study in healthy men. Neuropeptides 46, Cyranowski, J.M., Hofkens, T.L., Frank, E., Seltman, H., Cai, H.M., 211–215. Amico, J.A., 2008. Evidence of dysregulated peripheral Green, L., Fein, D., Modahl, C., Feinstein, C., Waterhouse, L., oxytocin release among depressed women. Psychosom. Med. Morris, M., 2001. Oxytocin and autistic disorder: alterations in 70, 967–975. peptide forms. Biol. Psychiatry 50, 609–613. De Dreu, C.K., Greer, L.L., Handgraaf, M.J., Shalvi, S., Van Kleef, Grewen, K.M., Girdler, S.S., Amico, J., Light, K.C., 2005. Effects of G.A., Baas, M., Ten Velden, F.S., Van Dijk, E., Feith, S.W., 2010. partner support on resting oxytocin, cortisol, norepinephrine, The neuropeptide oxytocin regulates parochial altruism in and blood pressure before and after warm partner contact. intergroup conflict among humans. Science 328, 1408–1411. Psychosom. Med. 67, 531–538. 76 brain research 1580 (2014) 69–77

Grippo, A.J., Gerena, D., Huang, J., Kumar, N., Shah, M., Ughreja, Kubzansky, L.D., Mendes, W.B., Appleton, A.A., Block, J., Adler, R., Carter, C.S., 2007. Social isolation induces behavioral and G.K., 2012. A heartfelt response: oxytocin effects on response neuroendocrine disturbances relevant to depression in female to social stress in men and women. Biol. Psychol. 90, 1–9. and male prairie voles. Psychoneuroendocrinology 32, Light, K.C., Grewen, K.M., Amico, J.A., 2005. More frequent partner 966–980. hugs and higher oxytocin levels are linked to lower blood Guastella, A.J., Mitchell, P.B., Dadds, M.R., 2008a. Oxytocin pressure and heart rate in premenopausal women. Biol. increases gaze to the eye region of human faces. Biol. Psychol. 69, 5–21. Psychiatry 63, 3–5. Lischke, A., Berger, C., Prehn, K., Heinrichs, M., Herpertz, S.C., Guastella, A.J., Mitchell, P.B., Mathews, F., 2008b. Oxytocin Domes, G., 2012a. Intranasal oxytocin enhances emotion enhances the encoding of positive social memories in recognition from dynamic facial expressions and leaves eye- humans. Biol. Psychiatry 64, 256–258. gaze unaffected. Psychoneuroendocrinology 37, 475–481. Guastella, A.J., Einfeld, S.L., Gray, K.M., Rinehart, N.J., Tonge, B.J., Lischke, A., Gamer, M., Berger, C., Grossmann, A., Hauenstein, K., Lambert, T.J., Hickie, I.B., 2010. Intranasal oxytocin improves Heinrichs, M., Herpertz, S.C., Domes, G., 2012b. Oxytocin emotion recognition for youth with autism spectrum increases amygdala reactivity to threatening scenes in disorders. Biol. Psychiatry 67, 692–694. females. Psychoneuroendocrinology 37, 1431–1438. Guastella, A.J., Hickie, I.B., McGuinness, M.M., Otis, M., Woods, MacDonald, E., Dadds, M.R., Brennan, J.L., Williams, K., Levy, F., E.A., Disinger, H.M., Chan, H.K., Chen, T.F., Banati, R.B., 2013. Cauchi, A.J., 2011. A review of safety, side-effects and Recommendations for the standardisation of oxytocin nasal subjective reactions to intranasal oxytocin in human administration and guidelines for its reporting in human research. Psychoneuroendocrinology 36, 1114–1126. research. Psychoneuroendocrinology 38, 612–625. Mantella, R.C., Vollmer, R.R., Li, X., Amico, J.A., 2003. Female Haxby, J.V., Hoffman, E.A., Gobbini, M.I., 2000. The distributed oxytocin-deficient mice display enhanced anxiety-related human neural system for face perception. Trends Cogn. Sci. 4, behavior. Endocrinology 144, 2291–2296. 223–233. Marsh, A.A., Yu, H.H., Pine, D.S., Blair, R.J., 2010. Oxytocin Heim, C., Young, L.J., Newport, D.J., Mletzko, T., Miller, A.H., improves specific recognition of positive facial expressions. Nemeroff, C.B., 2009. Lower CSF oxytocin concentrations in Psychopharmacology (Berl). 209, 225–232. women with a history of childhood abuse. Mol. Psychiatry 14, McCarthy, M.M., McDonald, C.H., Brooks, P.J., Goldman, D., 1996. 954–958. An anxiolytic action of oxytocin is enhanced by estrogen in Heinrichs, M., Meinlschmidt, G., Neumann, I., Wagner, S., the mouse. Physiol. Behav. 60, 1209–1215. Kirschbaum, C., Ehlert, U., Hellhammer, D.H., 2001. Effects of McEwen, B.B., 2004. Brain-fluid barriers: relevance for theoretical suckling on hypothalamic-pituitary-adrenal axis responses to controversies regarding vasopressin and oxytocin memory psychosocial stress in postpartum lactating women. J. Clin. research. Adv. Pharmacol. 50, 655–708. Endocrinol. Metab. 86, 4798–7804. Mens, W.B., Laczi, F., Tonnaer, J.A., de Kloet, E.R., van Wimersma Heinrichs, M., Baumgartner, T., Kirschbaum, C., Ehlert, U., 2003. Greidanus, T.B., 1983. Vasopressin and oxytocin content in Social support and oxytocin interact to suppress cortisol and cerebrospinal fluid and in various brain areas after subjective responses to psychosocial stress. Biol. Psychiatry administration of histamine and pentylenetetrazol. 54, 1389–1398. Pharmacol. Biochem. Behav. 19, 587–591. Hollander, E., Novotny, S., Hanratty, M., Yaffe, R., DeCaria, C.M., Mikolajczak, M., Pinon, N., Lane, A., de Timary, P., Luminet, O., Aronowitz, B.R., Mosovich, S., 2003. Oxytocin infusion reduces 2010. Oxytocin not only increases trust when money is at repetitive behaviors in adults with autistic and Asperger’s stake, but also when confidential information is in the disorders. Neuropsychopharmacology 28, 193–198. balance. Biol. Psychol. 85, 182–184. Hurlemann, R., Patin, A., Onur, O.A., Cohen, M.X., Baumgartner, T., Metzler, S., Dziobek, I., Gallinat, J., Wagner, M., Maier, W., Modahl, C., Green, L., Fein, D., Morris, M., Waterhouse, L., Kendrick, K.M., 2010. Oxytocin enhances amygdala- Feinstein, C., Levin, H., 1998. Plasma oxytocin levels in autistic dependent, socially reinforced learning and emotional children. Biol. Psychiatry 43, 270–277. empathy in humans. J. Neurosci. 30, 4999–5007. Moos, F., Freund-Mercier, M.J., Guerne, Y., Guerne, J.M., Stoeckel, Insel, T.R., Shapiro, L.E., 1992. Oxytocin receptor distribution M.E., Richard, P., 1984. Release of oxytocin and vasopressin by reflects social organization in monogamous and polygamous magnocellular nuclei in vitro: specific facilitatory effect of voles. Natl. Acad. Sci. U. S. A. 89, 5981–5985. oxytocin on its own release. J. Endocrinol. 102, 63–72. Insel, T.R., Fernald, R.D., 2004. How the brain processes social Neumann, I.D., Maloumby, R., Beiderbeck, D.I., Lukas, M., information: searching for the social brain. Annu. Rev. Landgraf, R., 2013. Increased brain and plasma oxytocin after Neurosci. 27, 697–722. nasal and peripheral administration in rats and mice. Ionescu, I.A., Dine, J., Yen, Y.C., Buell, D.R., Herrmann, L., Holsboer, Psychoneuroendocrinology 38, 1985–1993. F., Eder, M., Landgraf, R., Schmidt, U., 2012. Intranasally Nonaka, N., Farr, S.A., Kageyama, H., Shioda, S., Banks, W.A., 2008. administered neuropeptide S (NPS) exerts anxiolytic effects Delivery of galanin-like peptide to the brain: targeting with following internalization into NPS receptor-expressing neurons. intranasal delivery and cyclodextrins. J. Pharmacol. Exp. Ther. Neuropsychopharmacology 37, 1323–1337. 325, 513–519. Jones, P.M., Robinson, I.C., 1982. Differential clearance of Parker, K.J., Hoffman, C.L., Hyde, S.A., Cummings, C.S., neurophysin and neurohypophysial peptides from the Maestripieri, D., 2010. Effects of age on cerebrospinal fluid cerebrospinal fluid in conscious guinea pigs. oxytocin levels in free-ranging adult female and infant rhesus Neuroendocrinology 34, 297–302. macaques. Behav. Neurosci. 124, 428–433. Kawashima, R., Sugiura, M., Kato, T., Nakamura, A., Hatano, K., Pedersen, C.A., Prange Jr., A.J., 1979. Induction of maternal Ito, K., Fukuda, H., Kojima, S., Nakamura, K., 1999. The human behavior in virgin rats after intracerebroventricular amygdala plays an important role in gaze monitoring. A PET administration of oxytocin. Proc. Natl. Acad. Sci. U. S. A. 76, study. Brain 122 (Pt 4), 779–783. 6661–6665. Keri, S., Kiss, I., Kelemen, O., 2009. Sharing secrets: oxytocin and Prehn, K., Kazzer, P., Lischke, A., Heinrichs, M., Herpertz, S.C., trust in schizophrenia. Soc. Neurosci. 4, 287–293. Domes, G., 2013. Effects of intranasal oxytocin on pupil Kosfeld, M., Heinrichs, M., Zak, P.J., Fischbacher, U., Fehr, E., 2005. dilation indicate increased salience of socioaffective stimuli. Oxytocin increases trust in humans. Nature 435, 673–676. Psychophysiology 50, 528–537. brain research 1580 (2014) 69–77 77

Rimmele, U., Hediger, K., Heinrichs, M., Klaver, P., 2009. Oxytocin Taylor, S.E., Gonzaga, G.C., Klein, L.C., Hu, P., Greendale, G.A., makes a face in memory familiar. J. Neurosci. 29, 38–42. Seeman, T.E., 2006. Relation of oxytocin to psychological Ring, R.H., Malberg, J.E., Potestio, L., Ping, J., Boikess, S., Luo, B., stress responses and hypothalamic-pituitary-adrenocortical Schechter, L.E., Rizzo, S., Rahman, Z., Rosenzweig-Lipson, S., axis activity in older women. Psychosom. Med. 68, 238–245. 2006. Anxiolytic-like activity of oxytocin in male mice: Taylor, S.E., Saphire-Bernstein, S., Seeman, T.E., 2010. Are plasma behavioral and autonomic evidence, therapeutic implications. oxytocin in women and plasma vasopressin in men Psychopharmacology (Berl). 185, 218–225. biomarkers of distressed pair-bond relationships?. Psychol. Robinson, I.C., Clark, R.G., Fairhall, K.M., Jones, P.M., Parsons, J.A., Sci. 21, 3–7. 1982. Effects of anti-oxytocin serum in Brattleboro rats. Ann. Theodoridou, A., Rowe, A.C., Penton-Voak, I.S., Rogers, P.J., 2009. N. Y. Acad. Sci. 394, 285–290. Oxytocin and social perception: oxytocin increases perceived Robinson, I.C., Coombes, J.E., 1993. Neurohypophysial peptides in facial trustworthiness and attractiveness. Horm. Behav. 56, cerebrospinal fluid: an update. Ann. N. Y. Acad. Sci. 689, 128–132. 269–284. Thompson, R.R., George, K., Walton, J.C., Orr, S.P., Benson, J., 2006. Rosenblum, L.A., Smith, E.L., Altemus, M., Scharf, B.A., Owens, Sex-specific influences of vasopressin on human social M.J., Nemeroff, C.B., Gorman, J.M., Coplan, J.D., 2002. Differing communication. Proc. Natl. Acad. Sci. U. S. A. 103, 7889–7894. concentrations of corticotropin-releasing factor and oxytocin Thorne, R.G., Pronk, G.J., Padmanabhan, V., Frey 2nd, W.H., 2004. in the cerebrospinal fluid of bonnet and pigtail macaques. Delivery of insulin-like growth factor-I to the rat brain and Psychoneuroendocrinology 27, 651–660. spinal cord along olfactory and trigeminal pathways following Ross, H.E., Young, L.J., 2009. Oxytocin and the neural mechanisms intranasal administration. Neuroscience 127, 481–496. regulating social cognition and affiliative behavior. Front Toloczko, D.M., Young, L., Insel, T.R., 1997. Are there oxytocin Neuroendocrinol. 30, 534–547. receptors in the primate brain?. Ann. N. Y. Acad. Sci. 807, Savaskan, E., Ehrhardt, R., Schulz, A., Walter, M., Schachinger, H., 506–509. 2008. Post-learning intranasal oxytocin modulates human Unkelbach, C., Guastella, A.J., Forgas, J.P., 2008. Oxytocin memory for facial identity. Psychoneuroendocrinology 33, selectively facilitates recognition of positive sex and 368–374. relationship words. Psychol. Sci. 19, 1092–1094. Scantamburlo, G., Hansenne, M., Fuchs, S., Pitchot, W., Marechal, Veening, J.G., de Jong, T., Barendregt, H.P., 2010. Oxytocin- P., Pequeux, C., Ansseau, M., Legros, J.J., 2007. Plasma oxytocin messages via the cerebrospinal fluid: behavioral effects; a levels and anxiety in patients with major depression. review. Physiol. Behav. 101, 193–210. Psychoneuroendocrinology 32, 407–410. Viero, C., Shibuya, I., Kitamura, N., Verkhratsky, A., Fujihara, H., Scheele, D., Striepens, N., Gunturkun, O., Deutschlander, S., Katoh, A., Ueta, Y., Zingg, H.H., Chvatal, A., Sykova, E., Maier, W., Kendrick, K.M., Hurlemann, R., 2012. Oxytocin Dayanithi, G., 2010. REVIEW: Oxytocin: crossing the bridge modulates social distance between males and females. between basic science and pharmacotherapy. CNS Neurosci. J. Neurosci. 32, 16074–16079. Ther. 16, e138–156. Schneiderman, I., Zagoory-Sharon, O., Leckman, J.F., Feldman, R., Williams, J.R., Insel, T.R., Harbaugh, C.R., Carter, C.S., 1994. 2012. Oxytocin during the initial stages of romantic Oxytocin administered centrally facilitates formation of a attachment: relations to couples’ interactive reciprocity. partner preference in female prairie voles (Microtus Psychoneuroendocrinology 37, 1277–1285. ochrogaster). J. Neuroendocrinol. 6, 247–250. Seckl, J.R., Lightman, S.L., 1987. Diurnal rhythm of vasopressin Winslow, J.T., Noble, P.L., Lyons, C.K., Sterk, S.M., Insel, T.R., 2003. but not of oxytocin in the cerebrospinal fluid of the goat: lack Rearing effects on cerebrospinal fluid oxytocin concentration of association with plasma cortisol rhythm. J. Endocrinol. 114, and social buffering in rhesus monkeys. 477–482. Neuropsychopharmacology 28, 910–918. Shamay-Tsoory, S.G., Fischer, M., Dvash, J., Harari, H., Perach- Winston, J.S., Strange, B.A., O’Doherty, J., Dolan, R.J., 2002. Bloom, N., Levkovitz, Y., 2009. Intranasal Administration of Automatic and intentional brain responses during evaluation Oxytocin Increases Envy and Schadenfreude (Gloating). Biol. of trustworthiness of faces. Nat. Neurosci. 5, 277–283. Psychiatry 66, 864–870. Young, L.J., Huot, B., Nilsen, R., Wang, Z., Insel, T.R., 1996. Species Singer, T., Snozzi, R., Bird, G., Petrovic, P., Silani, G., Heinrichs, M., differences in central oxytocin receptor gene expression: Dolan, R.J., 2008. Effects of oxytocin and prosocial behavior on comparative analysis of promoter sequences. brain responses to direct and vicariously experienced pain. J. Neuroendocrinol. 8, 777–783. Emotion 8, 781–791. Zak, P.J., Kurzban, R., Matzner, W.T., 2005. Oxytocin is associated Smith, A.L., Freeman, S.M., Stehouwer, J.S., Inoue, K., Voll, R.J., with human trustworthiness. Horm. Behav. 48, 522–527. Young, L.J., Goodman, M.M., 2012. Synthesis and evaluation of Zak, P.J., Stanton, A.A., Ahmadi, S., 2007. Oxytocin increases C-11, F-18 and I-125 small molecule radioligands for detecting generosity in humans. PLoS One 2, e1128. oxytocin receptors. Bioorg. Med. Chem. 20, 2721–2738. brain research 1580 (2014) 188–198

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Intranasal oxytocin in the treatment of autism spectrum disorders: A review of literature and early safety and efficacy data in youth

Evdokia Anagnostoua,n, Latha Sooryab, Jessica Briana, Annie Dupuisc, Deepali Mankada, Sharon Smilea, Suma Jacobd aBloorview Research Institute, University of Toronto, 150 Kilgour Road, Toronto, ON, Canada M4G 1R8 bRush University Medical Center, Department of Psychiatry, 2150 West Harrison Street, Chicago, IL 606103, USA cThe Hospital for Sick Children, Dalla Lana School of Public Health, University of Toronto, 555 University Avenue, Toronto, Canada M5G 1X8 dUniversity of Minnesota, Department of Psychiatry & Pediatrics, 2101 6th Street SE, Minneapolis, MN 55455, USA article info abstract

Article history: Background: There is a paucity of treatments targeting core symptom domains in Autism Accepted 28 January 2014 Spectrum Disorder (ASD). Several animal models and research in typically developing Available online 5 February 2014 volunteers suggests that manipulation of the oxytocin system may have therapeutic fi Keywords: potential for the treatment of social de cits. We review the literature for oxytocin and ASD fi Autism and report on early dosing, safety and ef cacy data of multi-dose oxytocin on aspects of Oxytocin social cognition/function, as well as repetitive behaviors and co-occurring anxiety Clinical trial within ASD. Methods: Fifteen children and adolescents with verbal IQsZ70 were diagnosed fi Social cognition with ASD using the ADOS and the ADI-R. They participated in a modi ed maximum Social function tolerated dose study of intranasal oxytocin (Syntocinon). Data were modeled using repeated measures regression analysis controlling for week, dose, age, and sex. Results: Among 4 doses tested, the highest dose evaluated, 0.4 IU/kg/dose, was found to be well tolerated. No serious or severe adverse events were reported and adverse events reported/ observed were mild to moderate. Over 12 weeks of treatment, several measures of social cognition/function, repetitive behaviors and anxiety showed sensitivity to change with some measures suggesting maintenance of effect 3 months past discontinuation of intranasal oxytocin. Conclusions: This pilot study suggests that daily administration of intranasal oxytocin at 0.4 IU/kg/dose in children and adolescents with ASD is safe and has therapeutic potential. Larger studies are warranted. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2014 Elsevier B.V. All rights reserved.

nCorresponding author. E-mail address: [email protected] (E. Anagnostou). http://dx.doi.org/10.1016/j.brainres.2014.01.049 0006-8993 & 2014 Elsevier B.V. All rights reserved. brain research 1580 (2014) 188–198 189

neuronal firing (Ludwig and Leng, 2006). The mechanisms 1. Introduction by which central and peripheral release of oxytocin is coordinated remain poorly explained. Autism Spectrum Disorder (ASD) refers to a group of neuro- Central release of oxytocin and its closely related peptide, developmental disorders characterized by impairments in Arginine Vasopressin (AVP), are involved in aspects of social social communication, and repetitive behaviors and cognition and function including social recognition, social restricted interests (http://www.DSM5.org). The rates of ASD memory, affiliative behaviors, mother–infant and male– are on the rise with recent CDC numbers estimating the female pair-bond formation, separation distress, and other prevalence of the disorder at 1 in 88 children (CDC, 2008). aspects of social attachment, as well as the regulation of The social communication construct as it pertains to the stress response (Meyer-Lindenberg et al., 2011). Of note, ASD diagnosis has been debated extensively, even within oxytocin and AVP differ only by two amino acids, share versions of the DSM classification criteria, but there is agree- evolutionary history, have overlapping functions, influence ment that it includes difficulties in social-emotional recipro- each other's pathways or receptors throughout development city, including difficulties in forming or maintaining peer (Hirasawa et al., 2003; Landgraf and Neumann, 2004; relationships, deficits in nonverbal behaviors and in play Ragnauth et al., 2004) and as such they should be considered skills. Restricted interests and repetitive behaviors are con- together in biobehavioral contexts. ceptualized to include repetitive motor behaviors such as arm In animal models, oxytocin has been shown to play flapping, rocking, bouncing, and spinning behaviors, as well critical roles in social processing, recognition, and bonding, as higher-level compulsive-like behaviors including rigid and also to influence stereotyped behaviors such as exag- routines, ritualistic behaviors, and restrictive interests. gerated grooming (Carter, 1998; Insel et al., 1999; Winslow Despite the high burden of this disorder, there is no et al., 2003). In mammals, the Oxytocin Receptor (OXTR) is to date approved anywhere in the world for the expressed at higher levels in early development (Shapiro treatment of social deficits or repetitive behaviors associated Shapiro and Insel, 1989; Tribollet et al., 1989). Oxytocin with ASD. A paucity of molecular targets available for the knockout-mice have been shown to maintain olfaction and development of novel therapeutics related to the pathophy- cognitive performance, but suffer deficits in social recogni- siology of ASD has been postulated to contribute to this gap. tion that were recovered by intraventricular oxytocin (OXT), Specifically, that were developed for other disorders although not by AVP administration (Ferguson et al., 2000). with potentially overlapping phenotypes were tested in ASD, OXTR knockout mice emit fewer ultrasonic vocalizations assuming that phenotypic spectra would be associated with compared to the wild type, in response to social isolation, neurobiologic spectra. The approach has been useful for the experience deficits in social discrimination, and demon- treatment of associated symptoms of ASD (e.g. atypical strate more aggressive behavior (Takayanagi et al., 2005). and associated with ASD, Similarly, AVPR1A knock-out mice have been reported to McCracken et al., 2002), but has not produced a single agent exhibit social memory deficits (Bielsky and Young, 2004), shown to be effective for core symptoms. With the explosion and the expression of the receptor gene in the lateral of findings related to the biology of ASD and the biology of septum enhances social recognition (Bielsky et al., 2005). underlying core symptom domains associated with ASD, The pattern of AVPR1A receptor expression in the brain there is an opportunity for translational research to facilitate appears to be determined by variation in the length of a the development of novel therapeutics. microsatellite in the promoter region of the gene (Hammock Oxytocin is a nine-amino-acid peptide (nonapeptide), and Young, 2005). which is synthesized primarily in the paraventricular and supraoptic nucleus of the hypothalamus, and released into the bloodstream by axon terminals in the posterior pituitary. Peripheral release of oxytocin facilitates uterine contractions 1.1. Human studies of neuropeptide hormones during labor and milk letdown. Other peripheral targets of oxytocin include the kidneys and the pancreas. In addition to An explosion of studies has examined the effect of adminis- its peripheral role as a hormone, oxytocin is also widely tering a single dose on OXT and social cognition in humans. distributed throughout the Central Nervous System (CNS) and This work has been reviewed elsewhere extensively (e.g. functions as a neuromodulator. For example, oxytocin is Macdonald and Feifel, 2013; Guastella and MacLeod, 2012; released within the bed nucleus of the stria terminalis, the McCall and Singer, 2012; Kumsta and Heinrichs, 2013). Single spinal cord, the anterior commissural nucleus, and the dose studies in healthy volunteers have reported on medial amygdala. Oxytocin fibers are evident in a variety of increased trust (Kosfeld et al., 2005), empathic accuracy brain regions thought to be involved in social perception and (Domes et al., 2006, Guastella et al., 2009), time spent looking cognition, as well as emotion regulation, including the amyg- at eyes (Guastella et al., 2008), and face identity recognition dala and hippocampus and the ventral tegmental area of the memory (Savaskan et al., 2008; Rimelle et al., 2009). Attenua- midbrain (Gordon et al., 2011). Oxytocin receptors are also tion of amydgala activity has been documented with single widely distributed in the CNS, although their distribution is dose of OXT vs. placebo (Kirsch et al., 2005; Zink and Meyer- highly species specific (e.g. Donaldson and Young, 2008). Lindenberg, 2012; Domes et al., 2007). Such imaging studies Oxytocin tract studies, at least in voles, would suggest that suggest that manipulation of the oxytocin system may oxytocin release is not limited to the synaptic cleft, that produce circuitry modification that is relevant to social dendritic release occurs and that it is independent of deficits. 190 brain research 1580 (2014) 188–198

1.2. Oxytocin/Vasopressin and ASD increased trust, and time spent looking at eyes. However, single dose studies have limited ability to predict the ther- A number of researchers have hypothesized that OXT may be apeutic potential of administering intranasal oxytocin over a connected to autism given that repetitive behaviors and period of time and as such multi-dose studies are critical in deficits in social interaction are core features of the disorder, evaluating the compound's long-term therapeutic potential and that this neuropeptide is involved in the regulation of (Macdonald and Feifel, 2013). social cognition and some repetitive behaviors. Abnormalities in the neural pathway for OXT could account for many 1.2.4. Early multi-dose studies of intranasal oxytocin in ASD features of autism including the early onset, predominance A small number of studies designed to evaluate the effect of in males, genetic loading, and neuroanatomical abnormal- multi-dose intranasal oxytocin on core symptom domains in ities (Insel et al., 1999; Domes et al., 2007). ASD is available. In a single case report, Kosaka et al. (2012), reported on a 16 year old girl with ASD who received 8IU 1.2.1. Oxytocin plasma levels and ASD every day for 2 months. The authors reported that it was well Individuals with ASD have been reported to have lower than tolerated and that improvements were noted both in social average levels of blood OXT level in comparison to typically communication as well as irritability, based on clinician developing controls matched for age (Modahl et al., 1998; judgment and parent report on a standardized scale. Andari et al., 2010), although not universally, depending on Tachibana et al. (2013) reported on a case series of 8 male gender differences and assay methodology utilized (Miller youth (ages 10–14 years) who received a total of 6 months et al., 2013). Higher levels of oxytocin precursor peptides have exposure of oxytocin in the following manner: for the first 2 also been reported to be expressed in early ASD development months they received 8 IU per dose twice a day followed by 2 with subsequent decrease with age (Green et al., 2001). months of 16 IU per dose twice a day, and finally 2 months of 24 IU per dose. Before each step, a 1–2 week period of placebo 1.2.2. Oxytocin and vasopressin receptor genes (OXTR and was administered. Improvements were noted in social and AVPR1A) and ASD communication scores based on direct observation on a Accumulating studies are reporting that single nucleotide structured assessment, but not on parental reports of mala- polymorphisms of the OXTR gene are associated with ASD daptive behaviors. In addition, Dadds et al. (2013) evaluated a and related disorders (Wu et al., 2005; Jacob et al., 2007; 5-day intervention with 38 male youths with ASD (ages 7–16 Lerer et al., 2008; Yrigollen et al., 2008; Ebstein et al., 2012; Liu years). Boys were administered oxytocin (12 or 24 IU units et al., 2010; Wermter et al., 2010). However, most SNPs depending on weight) or placebo during parent-child inter- reported are outside the protein coding regions and their action training. No improvements were noted in emotion functional significance remains unknown. There has been a recognition, social skills and other behavioral domains com- recent report of a rare genetic variation of the OXTR gene pared to placebo, based on parent, clinician, or direct obser- within the protein coding region associated with ASD (Ma vation measures. Although the model of combining oxytocin et al., 2013). A heterozygous deletion of the OXTR gene has with a social learning activity is of great interest, this study is also been reported in a patient with ASD and family history of limited by several factors including the use of an experi- OCD (Gregory et al., 2009). In addition to variations in coding mental psychosocial intervention with unknown effects as a sequence, there is some early data to suggest potential ‘monotherapy’, as well as limited exposure (e.g. 5 days) to epigenetic modification related to the OXTR gene in ASD. treatments targeting core skills deficits. Our group also Gregory et al. (2009) reported increased methylation of the recently published a pilot randomized trial of intranasal OXTR gene promoter as compared to controls in two inde- oxytocin vs. placebo in 19 adults with ASD (Anagnostou pendent samples, including postmortem temporal cortex et al., 2012). Adults with ASD demonstrated improvements tissue from 8 ASD/control pairs. in empathic accuracy, lower order repetitive behaviors and quality of life on both self-report and experimental testing. 1.2.3. Single dose OXT studies in ASD There have been a growing number of OXT single dose 1.2.5. Rationale for intranasal administration studies in the last decade. Initially, an intravenous adminis- Oxytocin is metabolized in the gut by chymotrypsin and as tration of oxytocin vs. placebo over a 4 h period (Hollander such it cannot be administered orally. Despite its short half- et al., 2003, 2007) facilitated the retention of social cognition life in the blood, the intravenous formulation has been found in participants with ASD and produced significant reduction to produce behavioral effects (Hollander et al., 2003, 2007; in repetitive behaviors – i.e., needing to know, repeating, Ring et al., 2006), but it is too invasive to administer. One ordering, needing to tell/ask, self-injury, and touching. Sub- alternative is intranasal oxytocin; it is thought to be absorbed sequent single dose studies have employed the intranasal through the highly permeable nasal mucosa and, in the case formulation. Guastella et al. (2010) randomized 16 adoles- of the related peptide vasopressin, it has been shown to cross cents to a cross-over placebo-controlled study of a single dose the blood brain barrier (Born et al., 2002), and produce rising of intranasal oxytocin and reported significant improvements CNS levels for at least 4 h after intranasal administration, in empathic accuracy as measured by the Reading-the-Mind- mitigating concerns about the peripheral half-life of this in-the-Eyes task (RMET) with minimal adverse effects. Andari compound. It is also easy to self-administer. et al. (2010) randomized 13 adults with ASD to a single dose of Given that oxytocin is involved in the regulation of social intranasal oxytocin and reported stronger interactions with a communication and some repetitive behaviors, and based on cooperative partner during a computerized ball game, emerging pilot data, the authors received funding by the brain research 1580 (2014) 188–198 191

Department of Defense to conduct a series of studies of last dose (week 12) and was lost to follow-up for week 16 intranasal oxytocin in children and adolescents with ASD. assessments, due to study burden and non-efficacy. Two First, a modified Maximum-Tolerated-Dose (MTD) study was participants dropped out before first dose was administered agreed upon in collaboration with Health Canada to identify a and were replaced in the MTD escalation scheme. maximum dose for multi-dose studies in children up to a There was no effect of dose on measures of efficacy. maximum of 24 IU/per dose adjusted for weight. The study However, with only 3 children per group, the study was not also aimed at evaluating safety of multi-dose dosing in this designed and not powered to examine efficacy with each age group and identifying measures sensitive to change to be dose range. Several measures were insensitive to change or used in a follow-up randomized controlled trial in this not affected by oxytocin in our sample. These tasks included population. We report here the results of the MTD study. the DANVA, the Benton Facial Recognition Test, some subt- ests of the LFI battery, the Real and Apparent Test and the Second order Irony and Empathy tasks. For some, our participants experienced ceiling effects (e.g. DANVA) and for 2. Results some the participants experienced floor effects (e.g. 2nd order irony and empathy tasks). Several other measures of social Fifteen children and adolescents (11 male; 4 females; mean function, social cognition, repetitive behaviors and anxiety age: 13.8 (2.4) years), with a diagnosis of high-functioning did show improvements across 12 weeks of treatment autism or Asperger's Disorder (ADOS:Social Communication:10.33 þ (Table 2). These include the ABC-SW scale, SRS, functional (3.21)); (ADI :19.87 (4.76)); (ADI : 15.80 (3.67)); social communication communication and social skills subtests of the BASC, gen- (ADI 5.80 (1.97)) were recruited into the 16 week study repetitive eral and separation anxiety subscales of the CASI, both with twice daily dosing. Mean full scale IQ was 101.47 (22.60). repetitive behavior measures (CYBOCS and RBS-R), as well Of the 4 doses tested for 12 weeks duration (0.2, 0.26, 0.33 as measures of social recognition, empathic accuracy and and 0.4 IU/kg/dose), the 0.4 IU/kg/dose was the maximum theory of mind. Lastly, some of the measures of social tolerated dose, as no serious adverse events (SAEs) or severe cognition and function showed carry over effects 3 months adverse events were noted at any dose level. Adverse events after discontinuation of drug. were mild to moderate, and either expected or typically Ten of 15 participants were noted to be global responders associated with the disorder (Table 1). The WRAML revealed based on Clinical global Impression CGI-I-Global at week 12 no adverse events related to memory (Verbal Recall p 0.16, ¼ and 6/14 participants remained responders 3 months after Picture Memory Recognition p 0.92, Verbal recognition ¼ the last dose was administered. In terms of the CGI-I-Social, p 0.6). There were no clinically significant alterations in ¼ 7/15 were classified as responders at week 12 and 6/14 CBC, electrolytes, liver/renal function and osmolality, or remained responders at week 24. electrocardiograms. There were no discontinuations due to adverse events (Fig. 1). One adolescent dropped out after the

Table 1 – Safety (# events severity, if present in more than 10% of participants). 3. Discussion

Neuropsychiatric disorders The study is contributing to accumulating literature to Emotional lability (3 mild, 4 moderate) suggest potential efficacy and safety of multi-dose intranasal Irritability (6 mild) Headache (9 mild, 2 moderate) oxytocin in children and youth with ASD. Although this is not Migraine (3 mild) a conventional dose finding protocol, doses up to 0.4 IU/kg/ dose, given twice a day over 12 weeks produced no severe or Gastrointestinal disorders serious adverse events, and no metabolic or EKG abnormal- Abdominal discomfort (3 mild) fi Nausea (1 mild, 1 moderate) ities. Two thirds of the sample was classi ed as global responders, and almost half the sample was classified as and infestations responders in social function by the CGI at 12 weeks. Of Upper respiratory Infections (5 mild, 2 moderate) particular interest, a large percent of week 12 responders Metabolism and nutrition disorders maintained improvements 3 months after the end of the Decreased appetite (5 mild) study, especially in the social domain. Skin and subcutaneous tissue disorders A large number of measures were examined with respect Rash (3 mild) to their sensitivity to change in this pilot study. The reader Itchy nose (2 mild) should note that this study was not a randomized controlled fi General disorders trial, the proper method to con dently answer questions of Fatigue (4 mild, 1 moderate) efficacy, and as such the results on outcome measures should be interpreted with caution. There is a lack of data on Respiratory, thoracic, and mediastinal disorders instruments that are sensitive to change with treatment, Asthma attack (2 mild) especially within the domain of social function. Our study Tooth disorders was designed to determine instruments appropriate for Tooth sensitivity/pain (1 mild, 1 moderate) follow-up in a larger randomized controlled trial. Within this Serious adverse events 0 context, improvements were noted in the social domain on both parent/caregiver report as well as direct assessments, 192 brain research 1580 (2014) 188–198

CONSORT Flow Diagram

Enrollment Assessed for eligibility (phone

screened (n=120) Excluded (n=103)

Not meeting inclusion criteria

(n=62) Consented (n=17) Declined to participate (n=41) Discontinued prior to

medication n=2

Received IN-OXT (n=15)

Received allocated intervention (n=15)

Lost to follow-up (n=1)

Discontinued intervention (n=0)

Analysis

Analysed (n=15)

Excluded from analysis (n= 0)

Fig. 1 – Consort flow diagram.

Table 2 – Estimates of 0–12 and 0–24 week change for the 14 outcomes of 35 with a statistically significant (unadjusted) effect.

Variable 0–12 week (95%CL) p Value 0–24 week (95%CL) p Value

Social function ABC_SW 3.00 (0.05;5.96) 0.05 2.91 ( 1.98;7.80) 0.2 À Srs_total_tscore 9.8 (3.8;15.8) 0.002 8.2 ( 0.4;16.8) 0.06 À Basc_social skills tscore 2.84 ( 5.31; 0.38) 0.03 2.15 ( 5.42; 1.13) 0.2 À À À À À Basc_functional communication tscore 4.10 ( 6.23; 1.97) 0.0006 4.19( 7.29; 1.10) 0.01 À À À À À À Social cognition RMET_hard items 0.14 ( 0.82;0.53) 0.7 1.15 ( 1.99; 0.31) 0.009 À À À À À LFI_match_maker_ID 10.9 ( 19.3; 2.6) 0.01 7.8 ( 18.0;2.5) 0.1 À À À À À LFI same different faces 9.21( 15.26–3.15) 0.005 7.4 ( 15.5;0.6) 0.07 À À À À LFI same different houses 10.73 ( 18.5; 2.91) 0.009 10.2( 20.3; 0.2) 0.05 À À À À À À Irony_Empathy_first order 1.64 ( 3.02; 0.27) 0.02 2.31 ( 4.08; 0.54) 0.01 À À À À À À Strange Stories Task 1.36 ( 2.72; 0.008) 0.05 0.28 ( 2.00;1.44) 0.7 À À À À Anxiety ASI – general anxiety tscore 11.0 (4.0;18.0) 0.004 6.5 ( 2.5;15.6) 0.4 À ASI – separation anxiety tscore 7.5 (0.2;14.8) 0.04 5.7 ( 3.3;14.7) 0.2 À Repetitive behaviors RBS-R total 17.3 (6.3; 28.3) 0.003 6.0 ( 11.8;23.9) 0.5 À CYBOCS 2.93 (0.95;4.91) 0.004 2.52 (0.11;4.93) 0.04 suggesting some consistency across measures using different most likely a measure of global ASD severity, as well as the measurement approaches. social skills and functional communication domains of the Positive signals in the social domain were found on the BASC. Of interest, gains noted by the SRS and the functional ABC-SW, a measure of social withdrawal, the SRS total score, communication domain of the BASC were maintained 3 brain research 1580 (2014) 188–198 193

months post end of treatment, potentially suggesting a social learning effect. A similar pattern was noted in aspects of 4. Experimental procedures social perception/cognition. Improvements in face recogni- fi tion, and theory of mind were noted at 12 weeks. Some This was a modi ed Maximum Tolerated Dose (MTD), open persisted at 24 weeks and empathic accuracy improvements label trial of intranasal oxytocin in children and adolescents were most evident at 24 weeks, further supporting the with ASD. hypothesis already in the literature that oxytocin may facil- itate social learning. Of note, some non-specific improve- 4.1. Participants ments were noted in the “same-different” tasks of the LFI that were not specific to social stimuli and therefore cannot Children and adolescents, ages 10–17 inclusive, with a diag- be interpreted as social recognition gains. Most likely such nosis of an ASD, and verbal IQ equal to or above 70, were effects represent a nonspecific effect on attention and gen- enrolled in a 16 week study (12 weeks of medication exposure eral, categorical learning. Anxiolytic effects were pronounced plus four weeks of follow-up off medication). Participants during treatment, but disappeared with discontinuation of were recruited through the Holland Bloorview Kids Rehabili- treatment. Lastly improvements were noted in repetitive tation Hospital clinical database as well as presentations at behavior over the 12 weeks with some maintenance of effect local conferences and media. Diagnosis was established using 3 months later. a diagnostic interview to established DSM-IV criteria for an ASD, utilizing the Autism Diagnostic Observation Schedule (ADOS) (Lord et al., 2000) and the Autism Diagnostic 3.1. Limitations Interview-Revised (ADI-R) (Lord et al., 1994) that were con- ducted by research reliable administrators. All eligibility This is a small modified MTD study. As such, the sample assessments were completed prior to randomization into size is too small to view both safety and efficacy data the study. without caution. In addition, as the MTD design was For inclusion in this study, participants had to have base- modified in collaboration with regulatory agencies to not line severity score (Clinical Global Impression – Severity exceed the commonly used 24IU dose, it is not clear that CGI-S) of Z4 (moderately ill), be on stable pharmacological doses higher than that would not confer benefitwithin and or non-pharmacologic educational, behavioral, and/or acceptable safety parameters.Further,intheabsenceof dietary interventions (4 3 months prior to screening), and placebo control, adverse events associated with oxytocin have normal physical examination and laboratory test may be inflated. Lastly, given the sample size, interactions results. We excluded patients born prior to 35 weeks gesta- with concomitant medications (Table 3)couldnotbe tional age, patients with any primary psychiatric diagnosis examined. other than autism at screening, those with current severe neurological disease, including, but not limited to, epilepsy/ seizure disorder (except simple febrile seizures), severe 3.2. Conclusions movement disorder, tuberous sclerosis, fragile X, and any other known genetic syndromes, or known abnormal MRI/ Accumulating data suggest that oxytocin manipulation may structural lesion of the brain, pregnant female patients, have therapeutic potential in ASD. Early single dose studies sexually active female patients on hormonal birth control as well as case series and early multidose data support this and sexually active females who do not use two types of non- hypothesis. Our data extends this data set, adds to the safety hormonal birth control, patients with a medical condition profile of the multidose intranasal oxytocin administration in that might interfere with the conduct of the study, confound children and youth, establishes maximum tolerated dosage interpretation of the study results, or endanger their own within a range established by regulatory agencies and sug- well-being, those who are sensitive to Syntocinon or any gests several measures as potential candidates to be used in components of its formulation, those with HIV, HBV, HCV, follow-up randomized controlled trials in this population, hemophilia, abnormal blood pressure, drug abuse (as per already funded. DSM criteria), immunity disorder or severe depression and those unable to tolerate venipuncture procedures for blood sampling. All patients or their legal guardians signed the – Table 3 Concomitant medications. institutionally approved informed consent/assent according Medication (indication) Frequency (out of 15) to the Helsinki agreement and tri-council policy.

a (ADHD like symptoms) 6 4.2. Dose selection and MTD design Melatonin (insomnia) 4 Clonidine (insomnia) 1 fi fi Atomoxetine (ADHD like symptoms) 1 A modi ed dose nding method was used to determine Sertraline (anxiety) 1 safety among four dose levels. In adults, the dose with most Risperidone (impulse control) 2 evidence is 24 IU/dose. This is the equivalent of 0.4 IU/kg for a 60 kg person. Therefore, the maximum dose to be tested in a Methylphenidate (Concerta:1); Lisdexamfetamine (Vyvance):1; youth approved by Health Canada was 0.4 IU/kg/dose for a Methylphenidate (Ritalin):1; Methylphenidate (Biphentin):2; maximum of 24 IU/dose. This is different from a traditional Dextroamphetamine (Adderal):1. MTD design where the upward titration does not typically 194 brain research 1580 (2014) 188–198

stop unless the participants experience serious adverse administration and determine safety of first dose. Partici- events. Half the dose (0.2 IU/kg/dose) was the minimum dose pants were advised to administer one spray every 30 s until approved and 2 intermediate doses were also evaluated: 0.26 they have reached the total number of sprays directed. They and 0.33 IU/kg/dose. were told to sit upright, hold the bottle up right, insert the Dose-finding escalations were done in groups of three nozzle into one nostril and inhale gently through the nose patients. The following dose escalation rules were applied: while pushing down on the nozzle to activate the pump.

1. Three patients were studied at the first dose level (0.2 IU/ 4.5. Medications kg bid). 2. If none of these patients experienced Dose Limiting Oxytocin (Syntocinon, NOVARTIS, Switzerland) was adminis- Toxicity (DLT), then the dose was escalated to the next tered in the form of IN-OXT. higher level in the three subsequent patients. 3. If one of three patients had experienced DLT at the current 4.6. Safety assessments dose, then up to three more patients were to be accrued at the same level. (a) If none of these three additional Participants were seen every two weeks during the 12-week patients experienced DLT, then the dose were to be study and then again at 16 weeks, 4 weeks after last exposure escalated in subsequent patients. (b) If one or more of to drug, for Clinical Global Improvement (CGI-I) ratings, vital these additional patients experienced DLT, then patient signs and adverse event monitoring. Adverse events were entry at that dose level would be stopped, the MTD elicited using the SMURF (Greenhill et al., 2004) at every visit exceeded and dose escalation would be stopped. Up to and the Wide Range Assessment of Memory and Learning three more patients would be treated at the next lower (WRAML) (Sheslow and Adams, 2005) was done at baseline dose. If zero out of three patients experience DLT at a dose and end visits (weeks 12 and 24) to document any potential of 0.4 IU/kg, an additional three patients were to be treated adverse events related to memory, based on early animal at that dose. Using this escalation scheme, the probability model concerns (Engelmann et al., 1996). Safety blood work of escalating to the next level if the true proportion of and EKG were done every 4 weeks. adverse effects was 10%, 20%, 30%, 40%, or 50% was 91%, 71%, 49%, 31% and 17% respectively. Each participant only 4.7. Efficacy assessments received one of the preset doses. As noted above, the aim of this study was to establish maximum tolerated dose and safety for IN-OXT in this age group in ASD. In addition, we piloted several outcome Twice a day dosing (morning and afternoon) was chosen measures related to social cognition/function, repetitive given that CNS levels of vasopressin seem to rise for approxi- behaviors and co-occurring anxiety to select those with mately 4 h post intranasal administration (Born et al., 2002), documented sensitivity to change. These outcomes will be and in order to capture the most hours in which children and used in the follow-up randomized controlled trial in a larger youth are most likely to be involved in social interactions. sample with ASD and this study will also be funded by the Department of Defense. 4.3. Definition of toxicity for MTD design 4.7.1. Social cognition measures fi fi fi Based on existing data, it was dif cult to nd a de nition of All social cognition measures were administered at baseline toxicity for this compound. Although no standard toxicity and then week 12 and week 24. studies have been published for IN-OXT in children, the Side Diagnostic Analysis of Nonverbal Accuracy (DANVA-2) fi Effect (SE) pro le of IN-OXT in adults after decades of post (Baum and Nowicki, 1989). The DANVA-2, a measure of marketing use has been very benign. Standard toxicity pro- emotion recognition across multiple modalities with estab- tocols are typically based on Serious Adverse Events (SAEs) to lished reliability and validity for age 3–100, was administered determine the Maximum Tolerated Dose (MTD). In our case, to participants at baseline, week 12 and week 24. we decided to determine toxicity as the presence of SAEs or Let's Face It! Skills Battery (or Victoria/Yale Face Proces- severe adverse effects that are deemed by the investigators to sing Battery; VYFPB; Wolf et al., 2008) is a battery of compu- be likely or probably related to drug. terized face processing tasks. The full battery consists of 11 separate computer-administered tests. It assesses 2 broad 4.4. Dosing schedule domains involving (1) the perception of facial identity and (2) the perception of facial expression, with good split-half We selected morning and afternoon dosing to try to influence reliabilities (40.75). the greatest number of hours when youth are in settings with Benton Face Recognition Task (BFRT) (Benton et al., 1994) increased potential for social interaction (school, afterschool). is a standardized test frequently used to assess perceptual Moreover, this regimen was associated with improvements in skill in developmentally delayed populations, with good our adult pilot study (Anagnostou et al., 2012). Medication reliability. It is appropriate for participants aged 6–74 and was administered by the parents before school and right after provides a standardized and objective procedure for asses- school. All patients received their child's first dose by the sing the capacity to identify and discriminate photographs of study physician to educate parents and themselves on proper unfamiliar human faces. brain research 1580 (2014) 188–198 195

The Revised Eyes Test (Baron-Cohen et al., 2001) assesses Behavior Assessment System for Children (BASC) (Reynolds, the ability to decipher mental states of others. Participants 1992). We used the parent report scale with a focus on are presented with photographs of the eye-region of male functional communication and social skills subscales. and female actors accompanied by four descriptive words (e.g., “serious,”“ashamed,”“alarmed,” or “bewildered”) and 4.7.3. Repetitive behaviors must select the word that best describes what the actors are Both measures were administered at every visit thinking or feeling. In addition to restricting stimulus pre- sentation to the eye region, this test differs from others in  Child Yale-Brown Obsessive-Compulsive Scale (C-YBOCS) that it assesses more complex emotional states (e.g., shame, (Goodman et al., 1989, 65, modified by Scahill et al., 1997) bewilderment). The Yale-Brown Obsessive-Compulsive Scale is a Strange Stories Task (Happé, 1994) assesses the ability to clinician-rated questionnaire measuring the time spent, interpret nonliteral statements from stories read to the distress, interference, resistance, and control in relation participant. Questions probe whether the child understands to obsessions and compulsions based on a 5-point scale. that (a) a nonliteral statement has been made and (b) the The Compulsion subscale has been shown to be a reliable intent behind the statement (i.e., was the speaker was lying, and valid scale in ASD, and in measuring change in being sarcastic, or joking?). treatment studies of autism, and we used the “adapted Real and Apparent Emotion task: (Dennis et al., 2000) for PDD” version. evaluates children's understanding of real and deceptive fi  Repetitive Behavior Scale-Revised (RBS-R) (Bod sh et al., emotion in short narratives. The task, typically mastered 2000). The RBS-R is a parent-report measure developed to between the 6–8 years of age, represents an advanced capture the breadth of repetitive behaviors that are specific assessment of theory of mind in that it requires judgments to autism. It consists of 43-items that tap six repetitive about socially expressed emotion in the context of belief. behavior subtypes: Stereotyped, Self-injurious, Compulsive, Irony and Empathy Task (Dennis et al., 2001) measures Ritualistic, Sameness, and Restricted Interests. children's understanding of first and second order intention- ality associated with literal truth, ironic criticism, and decep- tive praise. Six everyday situations depict a task (e.g. tidying a room, baking a cake), with two participants, a speaker who 4.7.4. Anxiety measures makes a comment about the task and hearer who did the Child and Adolescent Symptom Inventory (CASI) (Gadow and task. The task has been well published in children with Sprafkin, 2009): The CASI is a clinician-reviewed, caregiver traumatic brain injury. report form based on the DSM-IV used to assess comorbid psychiatric symptoms. The anxiety subscales were used. 4.7.2. Social function measures Participants used a medication diary to mark down every Aberrant Behavior Checklist (ABC) (Aman et al., 1985) The time they took the medication. The diary was reviewed at ABC is a symptom checklist for assessing problem behaviors every visit and the study clinician and the participant (ages 6–54 years). The checklist consists of 58 items broken problem solved together in the case of missed doses to down into 5 subscales; only the Social Withdrawal subscale improve compliance. was used in this study. It was administered at baseline and weeks 12 and 24. 4.8. Statistical approach Social Responsiveness Scale (SRS) (Constantino, 2002) is a fi caregiver/educator rating scale of social behaviors speci c to Data were analyzed using SAS 9.3 (ref 2002–2010 by SAS ASD. It is a sensitive and reliable assessment that is designed Institute Inc., Cary, NC, USA). We plotted all outcomes across to aide diagnosis and measure treatment response with 4–18 time to visually assess any trends across dose. Data were years of age. The SRS measures social awareness, informa- modeled using repeated measures regression analysis con- tion processing, and social motivation and yields a quantita- trolling for week, dose, age, and sex. The changes from tive score that has been useful in endophenotyping studies of baseline to week 12 and baseline to week 24 and their 95% ASD and was administered at every visit. confidence limits were estimated from the model for each of CGI-I-Social: The CGI-I is a well validated measure the 35 outcomes variables examined. Eight of the outcomes employing a 7-point scale of clinical global impression of were also available as raw scores– in all cases, the results improvement (very much improved, much improved, mini- were similar using either score and only t-scores are pre- mally improved, no change, minimally worse, much worse, sented in the results. very much worse) that the clinician fills out after considering all the available information on the subject including the parent history, the examination in clinic, reports from the Competing interests/Financial disclosures school and other sources (Guy, 1976). In this case, the focus of the interview was on aspects of social function: approach, Dr. Evdokia Anagnostou has received consultation fees from turn taking, peer relationships and was administered at Seaside Therapeutics and Novartis and an unrestricted grant every visit. from Sanofi – Aventis Canada. Behavioral Assessment System for Children-2 is used to Dr. Latha Soorya, Dr. Jessica Brian, Dr. Annie Dupuis, Dr. evaluate the behavior and perceptions of children and young Deepali Mankad, and Dr. Sharon Smile have nothing to adults (ages 2–25 years). The BASC-2 is a revision of the disclose. Dr. Suma Jacob is a consultant to CogCubed games 196 brain research 1580 (2014) 188–198

and has as nothing to disclose with regards to pharmaceu- Benton, A.L., Sivan, A.B., KdeS, Hamsher, Varney, N.R., Spreen, O., tical companies. 1994. Contributions to Neuropsychological Assessment, 2nd edition Oxford University Press, New York. Bodfish, J.W., Symons, F.J., Parker, D.E., Lewis, M.H., 2000. Varieties of repetitive behavior in autism: comparisons to mental Funding retardation. J. Autism Dev. Disord. 30, 237–243. Born, J., Lange, T., Kern, W., McGregor, G.P., Bickel, U., Fehm, H.L., This study was funded by a Grant from the Department of 2002. Sniffing neuropeptides: a transnasal approach to the human brain. Nat. Neurosci. 5, 514–516. Defense (DOD) (AR093387) to Dr. Anagnostou. Carter, C.S., 1998. Neuroendocrine perspectives on social attachment and love. Psychoneuroendocrino 23, 779–818. CDC-Autism and Developmental Disabilities Monitoring Network Authors' contributions Surveillance Year; 2008 Principal Investigators; Centers for Disease Control and Prevention 30;61(3):1–19. PubMed PMID: 22456193. EA was the principal investigator for the study. She designed Constantino, J.N., 2002. The Social Responsiveness Scale. Western and ran the study, and prepared the manuscript. LS con- Psychological Services, Los Angeles. tributed to designing the study and manuscript preparation. Dadds, M.R., Macdonald, E., Cauchi, A., Williams, K., Levy, F., JB contributed to the design of the study, supervised psycho- Brennan, J., 2013. Nasal oxytocin for social deficits in logical assessments, and contributed to manuscript prepara- childhood autism: a randomized controlled trial. J. Autism tion. SJ contributed to designing the study, preparing the Dev. Disord.. Dennis, M., Lockyer, A., Lazenby, A., 2000. How high-functioning manuscript and is a site-PI on the subsequent RCT treatment children with autism understand real and deceptive emotion. trial. AD contributed to study design, data analysis and Autism 4, 370–381. manuscript preparation. DM and SS were critical to running Dennis, M., Purvis, K., Barnes, M.A., Wilkinson, M., Winner, E., the study and contributed to manuscript preparation. All 2001. Understanding of literal truth, ironic criticism, and authors read and approved the final manuscript. deceptive praise following childhood head injury. Brain Lang. 78, 1–16. Domes, G., Heinrichs, M., Gla¨scher, J., Bu¨ chel, C., Braus, D.F., Herpertz, S.C., 2007. Oxytocin attenuates amygdala responses Acknowledgments to emotional faces regardless of valence. Biol. Psychiatry 62, 1187–1190. The authors would like to thank Ms. Nadia Tanel, Ellen Domes, G., Heinrichs, M., Michel, A., Berger, C., Herpertz, S.C., Drumm, Lisa Genore and Dina Zaghloul as well as Mr. Brian 2006. Oxytocin improves “Mind-Reading” in humans. Biol. Psychiatry 61, 731–733. Crowe for their assistance in running the study and in the Donaldson, Z.R., Young, L.J., 2008. Oxytocin, vasopressin, and the preparation of this manuscript. neurogenetics of sociality. Science 322, 900–904. Ebstein, R.P., Knafo, A., Mankuta, D., Chew, S.H., Lai, P.S., 2012. references The contributions of oxytocin and vasopressin pathway genes to human behavior. Horm. Behav. 61, 359–379. Engelmann, M., Wotjak, C.T., Neumann, I., Ludwig, M., Landgraf, Aman, M.G., Singh, N.N., Stewart, A.W., Field, C.J., 1985. The R., 1996. Behavioral consequences of intracerebral vasopressin Aberrant Behavior Checklist: a behavior rating scale for the and oxytocin: focus on learning and memory. Neurosci. assessment of treatment effects. Am. J. Ment. Defic. 89, Biobehav. Rev. 20, 341–358. 485–491. Ferguson, J.N., Young, L.J., Hearn, E.F., Matzuk, M.M., Insel, T.R., Anagnostou, E., Soorya, L., Chaplin, W., Bartz, J., Halpern, D., Winslow, J.T., 2000. Social amnesia in mice lacking the Wasserman, S., Wang, A.T., Pepa, L., Tanel, N., Kushki, A., oxytocin gene. Nat. Genet. 25, 284–288. Hollander, E., 2012. Intranasal oxytocin versus placebo in Gadow, K.D., Sprafkin, J., 2009. The Symptom Inventories: An the treatment of adults with autism spectrum disorders: Annotated Bibliography. Checkmate Plus, Stony Brook, NY. a randomized controlled trial. Mol. Autism 3 (1), 16. Goodman, W.K., Price, L.H., Rasmussen, S.A., Mazure, C., Andari, E., Duhamel, J., Zalla, T., Herbrecht, E., Leboyer, M., Sirigu, Fleischmann, R.L., Hill, C.L., Heninger, G.R., Charney, D.S., A., 2010. Promoting social behavior with oxytocin in high 1989. The Yale-Brown obsessive compulsive scale. I. functioning autism spectrum disorders. Proc. Natl. Acad. Sci. Development, use, and reliability. Arch. Gen. Psychiatry 46, 107, 4389–4394. 1006–1011. Baron-Cohen, S., Wheelwright, S., Hill, J., Raste, Y., Plumb, I., 2001. Gordon, I., Martin, C., Feldman, R., Leckman, J.F., 2011. Oxytocin The “Reading the Mind in the Eyes” Test revised version: and social motivation. Dev. Cognit. Neurosci. 1, 471–493. a study with normal adults, and adults with Asperger Green, L.A., Fein, D., Modahl, C., Feinstein, C., Waterhouse, L., syndrome or high-functioning autism. J. Child Psychol. Morris, M., 2001. Oxytocin and autistic disorder: alteration in Psychiatry 42, 241–251. peptide forms. Biol. Psychiatry 50, 609–613. Baum, K.M., Nowicki, S., 1989. Perception of emotion: measuring Greenhill, L.L., Vitiello, B., Fisher, P., Levine, J., Davies, M., Abikoff, H., decoding accuracy of adult prosodic cues varying in intensity. Chrisman, A.K., Chuang, S., Findling, R.L., March, J., Scahill, L., J. Nonverbal Behav. 22, 89–107. Walkup, J., Riddle, M.A., 2004. Comparison of increasingly Bielsky, I.F., Hu, S.B., Young, L.J., 2005. in the detailed elicitation methods for the assessment of adverse vasopressin system: lack of an altered behavioral phenotype events in pediatric psychopharmacology. J. Am. Acad. Child in female V1a receptor knockout mice. Behav. Brain Res. 164, Adolesc. Psychiatry 43, 1488–1496. 132–136. Gregory, S., Connelly, J.J., Towers, A.J., Johnson, J., Biscocho, D., Bielsky, I.F., Young, L.J., 2004. Oxytocin, vasopressin, and social Markunas, C.A., Lintas, C., Abramson, K.A., Wright, H.H., Ellis, P., recognition in mammals. Peptides 25, 1565–1574. Langford, C.F., Worley, G., Delong, G.R., Murphy, S.K., brain research 1580 (2014) 188–198 197

Cuccaro, M.L., Persico, A., Pericak-Vance, M., 2009. Genomic and Liu, X., Kawamura, Y., Shimada, T., Otowa, T., Koishi, S., epigenetic evidence for oxytocin receptor deficiency in autism. Sugiyama, T., Nishida, H., Hashimoto, O., Nakagami, R., BMC Med. 7, 62. Tochigi, M., Umekage, T., Kano, Y., Miyagawa, T., Kato, N., Guastella, A., Einfeld, S., Gray, K., Rinehart, N., Tonge, B., Lambert, Tokunaga, K., Sasaki, T., 2010. Association of the oxytocin T.J., Hickie, I.B., 2009. Intranasal oxytocin improves emotion receptor (OXTR) gene polymorphisms with autism spectrum recognition for youth with autism spectrum disorders. Biol. disorder (ASD) in the Japanese population. J. Hum. Genet. 55, Psychiatry 67, 692–694. 137–141. Guastella, A.J., Mitchell, P.B., Dadds, M.R., 2008. Oxytocin Lord, C., Risi, S., Lambrecht, L., Cook, E.H., Leventhal, B.L., increases gaze to the eye region of human faces. Biol. DiLavore, P.C., Pickles, A., Rutter, M., 2000. The autism Psychiatry 63, 3–5. diagnostic observation schedule-generic: a standard measure Guastella, A.J., Einfeld, S.L., Gray, K.M., Rinehart, N.J., Tonge, B.J., of social and communication deficits associated with the Lambert, T.J., Hickie, I.B., 2010. Intranasal oxytocin improves spectrum of autism. J. Autism Dev. Disord. 30, 205–223. emotion recognition for youth with autism spectrum Lord, C., Rutter, M., Le Couteur, A., 1994. Autism Diagnostic disorders. Biol. Psychiatry 67, 692–694. Interview-Revised: a revised version of a diagnostic interview Guastella, A.J., MacLeod, C., 2012. A critical review of the for caregivers of individuals with possible pervasive influence of oxytocin nasal spray on social cognition in developmental disorders. J. Autism Dev. Disord. 24, 659–685. humans: evidence and future directions. Horm. Behav. 61, Ludwig, M., Leng, G., 2006. Dendritic peptide release and 410–418. peptidedependent behaviors. Nat. Rev. Neurosci. 7, 126–136. Guy, W., 1976. ECDEU Assessment Manual for Ma, W.J., Hashii, M., Munesue, T., Hayashi, K., Yagi, K., Yamagishi, M., psychopharmacology, revised. National Institute of Mental Higashida, H., Yokoyama, S., 2013. Non-synonymous single- Health, Rockville, MD. nucleotide variations of the human oxytocin receptor gene Hammock, E.A., Young, L.J., 2005. Microsatellite instability and autism spectrum disorders: a case-control study in a generates diversity in brain and sociobehavioral traits. Japanese population and functional analysis. Mol. Autism 4, Science 308, 1630–1634. 22. Happe´, F.G., 1994. An advanced test of theory of mind: Macdonald, K., Feifel, D., 2013. Helping oxytocin deliver: understanding of story characters’ thoughts and feelings by considerations in the development of oxytocin-based able autistic, mentally handicapped, and normal children and therapeutics for brain disorders. Front. Neurosci. 7, 35. adults. J. Autism Dev. Disord. 24, 129–154. McCall, C., Singer, T., 2012. The animal and human Hirasawa, M., Mouginot, D., Kozoriz, M.G., Kombian, S.B., Pittman, neuroendocrinology of social cognition, motivation and Q.J., 2003. Vasopressin differentially modulates non-NMDA behavior. Nat. Neurosci. 15, 681–688. receptors in vasopressin and oxytocin neurons in the McCracken, J.T., McGough, J., Shah, B., Cronin, P., Hong, D., Aman, supraoptic nucleus. J. Neurosci. 23, 4270–4277. M.G., McMahon, D., 2002. Risperidone in children with autism Hollander, E., Bartz, J., Chaplin, W., Phillips, A., Sumner, J., Soorya, L., and serious behavioral problems. N. Engl. J. Med. 347, 314–321. Anagnostou, E., Wasserman, S., 2007. Oxytocin increases Meyer-Lindenberg, A., Domes, G., Kirsch, P., Heinrichs, M., 2011. retention of social cognition in autism. Biol. Psychiatry 61, Oxytocin and vasopressin in the human brain: social 498–503. neuropeptides for translational medicine. Nat. Rev. Neurosci. Hollander, E., Novotny, S., Hanratty, M., Yaffe, R., DeCaria, C.M., 19 (12), 524–538. Aronowitz, B.R., Mosovich, S., 2003. Oxytocin infusion reduces Miller, M., Bales, K.L., Taylor, S.L., Yoon, J., Hostetler, C.M., Carter, repetitive behaviors in adults with autistic and Asperger’s C.S., Solomon, M., 2013. Oxytocin and vasopressin in children disorders. Neuropsychopharmacology 28, 193–198. and adolescents with autism spectrum disorders: sex Insel, T.R., Obrien, D.J., Leckman, J.F., 1999. Oxytocin, vasopressin, differences and associations with symptoms. Autism Res. 6, and autism: is there a connection. Biol. Psychiatry 45, 145–157. 91–102. Jacob, S., Brune, C.W., Carter, C.S., Leventhal, B.L., Lord, C., Cook, Modahl, C., Green, L.A., Fein, D., Morris, M., Waterhouse, L., E.H., 2007. Association of the oxytocin receptor gene (OXTR) in Feinstein, C., Levin, H., 1998. Plasma oxytocin levels in autistic Caucasian children and adolescents with autism. Neurosci. children. Biol. Psychiatry 43, 270–277. Lett. 417, 6–9. Ragnauth, A.K., Goodwillie, A., Brewer, C., Muglia, L.J., Pfaff, D.W., Kirsch, P., Esslinger, C., Chen, Q., Mier, D., Lis, S., Siddhanti, S., Kow, L.M., 2004. Vasopressin stimulates ventromedial Gruppe, H., Mattay, V.S., Gallhofer, B., Meyer-Lindenberg, A., hypothalamic neurons via oxytocin receptors in oxytocin 2005. Oxytocin modulates neural circuitry for social cognition gene knockout male and female mice. Neuroendocrinology 80, and fear in humans. J. Neurosci. 25, 11489–11493. 92–99. Kosaka, H., Munesue, T., Ishitobi, M., Asano, M., Omori, M., Sato, Reynolds Kamphaus, Behavior Assessment System for Children. M., Tomoda, A., Wada, Y., 2012. Long-term oxytocin Pearson Assessment, 1992. administration improves social behaviors in a girl with Rimelle, U., Hediger, K., Heinrichs, M., Klaver, P., 2009. Oxytocin autistic disorder. BMC Psychiatry 12, 110. makes a face memory familiar. J. Neurosci. 29, 38–42. Kosfeld, M., Heinrichs, M., Zak, P.J., Fischbacher, U., Fehr, E., 2005. Ring, R.H., Malberg, J.E., Potestio, L., Ping, J., Boikess, S., Luo, B., Oxytocin increases trust in humans. Nature 435, 673–676. Schechter, L.E., Rizzo, S., Rahman, Z., 2006. Rosenzweig- Kumsta, R., Heinrichs, M., 2013. Oxytocin, stress and social Lipson Sl. Anxiolytic-like activity of oxytocin in male mice: behavior: neurogenetics of the human oxytocin system. Curr. behavioral and autonomic evidence, therapeutic implications. Opin. Neurobiol. 23, 11–16. Psychopharmacology 185 (2), 218–225. Landgraf, R., Neumann, I.D., 2004. Vasopressin and oxytocin Savaskan, E., Ehrhardt, R., Schulz, A., Walter, M., Schachinger, H., release within the brain: a dynamic concept of multiple and 2008. Post-learning intranasal oxytocin modulates human variable modes of neuropeptide communication. Front memory for facial identity. Psychoneuroendocrinology 33, Neuroendocrinol. 25, 150–176. 368–374. Lerer, E., Levi, S., Salomon, S., Darvasi, A., Yirmiya, N., Ebstein, R. Scahill, L., Riddle, M.A., McSwiggin-Hardin, M., Ort, S.I., King, R.A., P., 2008. Association between the oxytocin receptor (OXTR) Goodman, W.K., Cicchetti, D., Leckman, J.F., 1997. Children’s gene and autism: relationship to Vineland Adaptive Behavior Yale-Brown Obsessive Compulsive Scale: reliability and validity. Scales and cognition. Mol. Psychiatry 13, 980–988. J. Am. Acad. Child Adolesc. Psychiatry 36, 844–852. 198 brain research 1580 (2014) 188–198

Sheslow D., Adams W. Wide Range Assessment of Memory and involvement of genetic variation in the oxytocin receptor gene Learning, Second Edition (WRAML-2). PAR, 2005. (OXTR) in the etiology of autistic disorders on high- Shapiro Shapiro, L.E., Insel, T.R., 1989. Ontogeny of oxytocin functioning level. Am. J. Med. Genet. B Neuropsychiatr. Genet. receptors in rat forebrain: a quantitative study. Synapse 4, 153B, 629–639. 259–266. Wolf, M.J., Tanaka, James W., Klaiman, Cheryl, Cockburn, Jeff, Takayanagi, Y., Yoshida, M., Bielsky, I.F., et al., 2005. Pervasive Herlihy, Lauren, Brown, Carla, South, Mikle, McPartland, social deficits, but normal parturition, in oxytocin receptor- James, Kaiser, Martha D., Phillips, Rebecca, Schultz, Robert T., deficient mice. Proc. Natl. Acad. Sci. USA 102, 16096–16101. 2008. Specific impairment of face-processing abilities in Tachibana, M., Kagitani-Shimono, K., Mohri, I., Yamamoto, T., children with autism spectrum disorder using the let’s face it! Sanefuji, W., Nakamura, A., Oishi, M., Kimura, T., Onaka, T., skills battery. Autism Res. 1, 329–340. Ozono, K., Taniike, M., 2013. Long-term administration of Wu, S., Jia, M., Ruan, Y., Liu, J., Guo, Y., Shuang, M., Gong, X., intranasal oxytocin is a safe and promising therapy for early Zhang, Y., Yang, X., Zhang, D., 2005. Positive association of the adolescent boys with autism spectrum disorders. J. Child oxytocin receptor gene (OXTR) with autism in the Chinese Adolesc. Psychopharmacol. 23, 123–127. Han population. Biol. Psychiatry 58, 74–77. Tribollet, E., Charpak, S., Schmidt, A., Dubois-Dauphin, M., Yrigollen, C.M., Han, S.S., Kochetkova, A., Babitz, T., Chang, J.T., Dreifuss, J.J., 1989. Appearance and transient expression of Volkmar, F.R., Leckman, J.F., Grigorenko, E.L., 2008. Genes oxytocin receptors in fetal, infant, and peripubertal rat brain controlling affiliative behavior as candidate genes for autism. studied by autoradiography and electrophysiology. J. Neurosci. Biol. Psychiatry 63, 911–916. 9, 1764–1773. Zink, C.F., Meyer-Lindenberg, A., 2012. Human neuroimaging of Winslow, J.T., Noble, P.L., Lyons, C.K., Sterk, S.M., Insel, T.R., 2003. oxytocin and vasopressin in social cognition. Horm. Behav. 61, Rearing effects on cerebrospinal fluid oxytocin concentration 400–409. and social buffering in rhesus monkeys. Neuropsychopharmacology 28, 910–918. Wermter, A.K., Kamp-Becker, I., Hesse, P., Schulte-Ko¨ rne, G., Strauch, K., Remschmidt, H., 2010. Evidence for the brain research 1580 (2014) 133–142

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Maternal oxytocin response predicts mother-to-infant gaze Sohye Kima,b,c, Peter Fonagyb,d, Orsolya Koose, Kimberly Dorsettb,1, Lane Strathearna,b,c,f,n aDepartment of Pediatrics, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA bMenninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA cAttachment and Neurodevelopment Laboratory, Children's Nutrition Research Center, 1100 Bates Street, Suite 4004, Houston, TX 77030, USA dResearch Department of Clinical, Educational and Health Psychology, University College London, 1-19 Torrington Place, London WC1E 7HB, United Kingdom eDepartment of Developmental Research, Institute for Psychology, Hungarian Academy of Sciences, 1394 Budapest, P.O. Box 398, Hungary fThe Meyer Center for Developmental Pediatrics, Texas Children'sHospital,8080N.StadiumDrive,Houston,TX77054,USA article info abstract

Article history: The neuropeptide oxytocin is importantly implicated in the emergence and maintenance of Accepted 27 October 2013 maternal behavior that forms the basis of the mother–infant bond. However, no research has yet Available online 1 November 2013 examined the specific association between maternal oxytocin and maternal gaze, a key modality Keywords: through which the mother makes social contact and engages with her infant. Furthermore, prior Oxytocin oxytocin studies have assessed maternal engagement primarily during episodes free of infant Mothering distress, while maternal engagement during infant distress is considered to be uniquely relevant – Mother-to-infant gaze to the formation of secure mother infant attachment. Two patterns of maternal gaze, maternal Distress gaze toward and gaze shifts away from theinfant,weremicro-coded while 50 mothers interacted fi Engagement with their 7-month-old infants during a modi ed still-face procedure. Maternal oxytocin response was defined as a change from baseline in the mother's plasma oxytocin level following interaction with her infant. The mother's oxytocin response was positively associated with the duration of time her gaze was directed toward her infant, while negatively associated with the frequency with which her gaze shifted away from her infant. Importantly, mothers who showed low/average oxytocin response demonstrated a significant decrease in their infant gaze during periods of infant distress, while such change was not observed in mothers with high oxytocin response. The findings underscore the involvement of oxytocin in regulating the mother's responsive engagement with her infant, particularly in times when the infant'sneed for access to the mother is greatest. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2013 Elsevier B.V. All rights reserved.

nCorresponding author at: Attachment and Neurodevelopment Laboratory, Department of Pediatrics, Children's Nutrition Research Center, Baylor College of Medicine/Texas Children's Hospital, 1100 Bates St, Suite 4004-B, Houston, TX 77030, USA. E-mail address: [email protected] (L. Strathearn). 1The author is now at Child Trauma Training Center, Department of Psychiatry, University of Massachusetts Medical School, 76 Church Street, Suite 301, Whitinsville, MA 01588.

0006-8993/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.brainres.2013.10.050 134 brain research 1580 (2014) 133–142

mothers' gaze (Stern, 1974)andbegintojoininmutualgaze 1. Introduction with their mothers as early as 3 months of age, which serves as abasisforthemother–infant synchrony that subsequently While the neuropeptide oxytocin has been recognized for its emerges in other modalities (e.g., touch, vocalization, facial functions in parturition and milk ejection for many decades, expression; Colonnesi et al., 2012; Feldman, 2007; Lavelli and there exists now a substantial literature underscoring the role Fogel, 2013; Tronick et al., 1980). of oxytocin in regulating social behaviors (see reviews, Despite its significance, the specific association between Benarroch, 2013; Feldman, 2012; Meyer-Lindenberg et al., 2011). maternal gaze and maternal oxytocin has not yet been Asizablenumberofstudieshaveimplicatedoxytocininmater- examined. Previous studies have measured maternal gaze, nal care (Febo et al., 2005; Pedersen et al., 2006; Strathearn et al., but only as part of a composite of maternal behaviors 2009b), pair bonding (Ross et al., 2009; Schneiderman et al., 2012), (also encompassing touch, vocalization, and affect; e.g., interpersonal trust (Van IJzendoorn and Bakermans-Kranenburg, Atzil et al., 2011; Feldman et al., 2007, 2010b; Gordon et al., 2012), emotion recognition (Lischke et al., 2012; Perry et al., 2013), 2010). Furthermore, while attachment literature has under- and empathy (Hurlemann et al., 2010; Rodrigues et al., 2009). scored that maternal sensitivity to infant distress uniquely Oxytocin has been characterized as a “hormone of affiliation” contributes to optimal socioemotional outcomes in the child (Insel, 1992)andtheoxytocinergicsystemhasreceivedattention (Leerkes et al., 2009; McElwain and Booth-LaForce, 2006), prior as a key neural substrate of maternal caregiving, involved in the studies have examined maternal oxytocin only in relation to emergence and maintenance of maternal behaviors (Feldman, indices of maternal synchrony during episodes of positive 2012; Strathearn, 2011). Many important advances in this regard affect and have not considered episodes of distress (Atzil have come from animal models (Francis et al., 2002; Keverne et al., 2011; Feldman et al., 2010b, 2011). and Kendrick, 1992; Maestripieri et al., 2009; Pedersen et al., 2006; In our previous functional magnetic resonance imaging Williams et al., 2001), and they have been extended to human (fMRI) study (Strathearn et al., 2009b), we demonstrated that subjects over the past decade, revealing both parallel and mothers' peripheral oxytocin responses predicted their blood divergent findings. oxygenation level-dependent (BOLD) brain responses to their In humans, peripheral oxytocin levels are higher in pregnant own infants' faces. The greater the mothers' oxytocin and parturient women than non-pregnant women (Feldman responses during interactions with their infants, the greater et al., 2007; Gordon et al., 2008). Oxytocin levels show high intra- was their activation in regions known to be rich in oxytocin individual stability over the course of pregnancy (Feldman et al., receptors (i.e., hypothalamic/pituitary region) when viewing 2007; Levine et al., 2007)andearlymotherhood(Gordon et al., their infants' faces in the scanner. In the present study, we 2010), suggesting that they may constitute a trait-like charac- extended our line of investigation to the mother's actual gaze teristic that underpins the expression of maternal behavior. behavior during real-time interaction with her infant. We Prospective and cross-sectionalstudieshavedemonstratedthat sought to examine the relationship between maternal oxyto- maternal oxytocin levels are systematically associated with cin response and mother-to-infant gaze during periods of naturally occurring variations in maternal behavior, with high infant non-distress as well as distress. plasma oxytocin levels during pregnancy and postpartum Mother–infant dyads were observed during 50-min semi- predicting increased maternalbehaviorinthepostpartum structured interaction sessions. Maternal oxytocin response months (Atzil et al., 2011; Feldman et al., 2007; Gordon et al., was defined as any change in the mother's oxytocin level 2010). Interaction with their young in the postpartum period following interaction with her infant compared to baseline. further stimulates oxytocin response in mothers (Feldman Two patterns of maternal gaze, duration of maternal gaze et al., 2010a, 2010b), though significant inter-individual varia- toward and frequency of gaze shifts away from the infant, tions have been found (Strathearn et al., 2012), as with the were coded during a well-validated interaction paradigm, baseline oxytocin levels. These natural variations in maternal a modified still-face procedure (MSFP; Koos and Gergely, oxytocin response have systematically predicted differences in 2001). The MSFP is a three-phase procedure, during which the quality of maternal care provided by mothers (Feldman the mother interacts freely with the infant in phases 1 and et al., 2010a, 2010b). 3, but is instructed to maintain a neutral ‘still face’ during Quality provision of maternal care and formation of secure phase 2, suddenly depriving the infant of maternal contin- attachment bonds are of particular importance in the early gency and inducing stress in the infant (Koos and Gergely, postpartum months, given their long-term effects on the 2001; Tronick et al., 1978; Fig. 1). The experimental manipula- development of the offspring (Fonagy et al., 2007; Kochanska tion reliably produces changes in the infant's level of distress: and Kim, 2013; Sroufe et al., 2005; Weinfield et al., 2004). It is infants display clear signs of distress during phase 2, which well established that sensitive and responsive maternal beha- have been shown to carry over to phase 3 (Haley and vior has direct bearings on the child's life-long capacity for Stansbury, 2003; Mesman et al., 2009). The MSFP thereby social adaptation and stress regulation (Kochanska, 2001; offers an opportunity to examine the mother's behavior in Mayes, 2006; Mikulincer and Shaver, 2007; Schore, 2001). One the absence and presence of signals of infant distress. We important channel through which maternal sensitive respon- hypothesized that maternal oxytocin response would be posi- siveness is communicated is mother-to-infant gaze. Gaze is a tively associated with maternal gaze toward the infant, and central modality through which mothers signal their availabil- negatively associated with maternal gaze shifts away from the ity, establish mutual engagement, and initiate regulation of infant. We further predicted that this association would become infant arousal, particularly in times of infant distress (Beebe more pronounced during periods of infant distress than non- et al., 2010; Slee, 1984). Infants are highly sensitive to their distress, in phase 3 compared with phase 1 of the MSFP. brain research 1580 (2014) 133–142 135

Table 1 – Sociodemographic and behavioral characteris- tics of mothers and infants (N 50). Dividing screen ¼ Characteristics Value Range Mother Infant Baseline OT, pg/ml a Mean7SD 1.757.89 .50–4.80 OT response, pg/ml b Mean7SD .067.93 3.70–3.15 À À Maternal age, years Mean7SD 28.074.6 19–41 Infant age, months Mean7SD 6.471.7 4–11 One-way mirror Infant sex, n (%) Male 21 (42.0) Female 29 (58.0) Marital status, n (%) Digital Married 38 (76.0) video cameras Not married 12 (24.0) Maternal race, n (%) White 31 (62.0) Non-White 19 (38.0) Maternal education, n (%) College incomplete 11 (22.0) Phase 1: College/university degree 22 (44.0) Normal interaction Postgraduate degree 16 (32.0) Socioeconomic status c Mean7SD 44.5715.4 15–66 Maternal IQ d Mean7SD 110.278.4 81–120 Phase 2: Maternal depression (BDI) e Still-face Mean7SD 5.374.8 0–19 Maternal personality pathology (PDQ) f Mean7SD 19.4711.5 3–50 Maternal parenting stress (PSI) Phase 3: Child Domain, Mean7SD 91.3714.1 64–121 Parent Domain, Mean7SD 112.8723.6 66–157 Recovery Total Stress, Mean7SD g 201.5734.5 109–264 Breastfeeding status h, n (%) Not breastfeeding 13 (26.0) Still breastfeeding 28 (56.0) Fig. 1 – The modified still-face procedure (MSFP): (a) diagram Daycare status i, n (%) of the experimental setting and (b) example still frames from Less than 20 h per week 20 (40.0) all three phases. More than 20 h per week 19 (38.0)

Note. BDI Beck Depression Inventory-II; PDQ Personality Disor- ¼ ¼ der Questionnaire 4 ; PSI Parenting Stress Index. À þ ¼ a Plasma oxytocin concentration was measured following a 20- min period of mother–infant separation (OT 1). 2. Results b Change in plasma oxytocin concentration between baseline (OT 1) and post mother–infant interaction (mean of OT 2 2.1. Participant characteristics and preliminary analyses and OT 3). c Socioeconomic status was estimated using Hollingshead (1975)'s Participant characteristics are shown in Table 1.Participants Four-Factor Index of Social Status, and represents joint informa- tion with partner, when applicable. were a non-clinical sample of first-time mothers who were d Maternal Full Scale IQ was estimated from the Wechsler Test of generally of middle to high socioeconomic status, three- Adult Reading (WTAR). quarters of whom held a college or postgraduate degree. All e BDI-II score of r9 indicates minimal depression. mothers scored below the clinical range for personality disorders, f PDQ-4 total score of Z50 is highly suggestive of DSM-IV þ while four mothers scored in the mildly depressed range, one of personality disorder. g whom also scored in the clinically significant range for parenting PSI total stress score of o260 is considered in the normal range. h stress. However, these four mothers did not differ from the rest Data were missing for 9 participants. i Hours per week that someone other than the mother looked after of the sample in oxytocin levels (ps4.65) or maternal gaze the infant. Data were missing for 11 participants. (ps4.20); neither did the exclusion of the four mothers signifi- cantly alter the results reported below. Maternal clinical char- acteristics are therefore not considered further. Maternal depression was unrelated to baseline oxytocin level (r and r .05, p .77), or infant affect variables BDI & baseline BDI & gaze shift away ¼ ¼ .01, p .95), oxytocin response (OTResp; r .09, (r .08, p .59 and r .04, OT ¼ ¼ BDI & OTResp ¼ BDI & positive affect ¼À ¼ BDI & negative affect¼À p .57), maternal gaze variables (r .08, p .62 p .78). ¼ BDI & gaze toward¼ ¼ ¼ 136 brain research 1580 (2014) 133–142

Breastfeeding status also did not correlate with oxytocin r .01, p .93; r progesterone & gaze toward ¼ ¼ estradiol & gaze shift away ¼ response (r .15, p .34), or with maternal .18, p .27 and r .13, p .38). breastfeeding & OTResp ¼ ¼ ¼ progesterone & gaze shift away ¼ ¼ gaze variables (r .10, p .54 and breastfeeding & gaze toward ¼ ¼ r .17, p .28). Measures of men- breastfeeding & gaze shift away ¼À ¼ strual cycle (i.e., estradiol and progesterone levels) were also 2.2. Infant affect during modified still-face procedure not associated with OTResp (restradiol & OTResp .07, p .65 ¼ ¼ As expected, and consistent with previous research, signifi- and rprogesterone & OTResp .06, p .67), or with maternal ¼ ¼ cant changes were noted in infant affect across the three gaze variables (restradiol & gaze toward .01, p .95 and ¼À ¼ phases of the MSFP (positive affect, F(2, 98) 33.78, po.001; ¼ negative affect, F(2, 98) 35.24, po.001; Fig. 2). Compared to ¼ baseline (phase 1), infants displayed increased negative affect during still-face (phase 2) (M .22, SD .26; M .51, phase 1 ¼ ¼ phase 2 ¼ SD .41; t – (49) 7.09, po.001), as well as decreased ¼ phases 2 1 ¼ positive affect (M .27, SD .24; M .04, SD .09; phase 1 ¼ ¼ phase 2 ¼ ¼ tphases 2 1(49) 7.83, po.001). While there was a rebound of À ¼À positive affect during recovery (phase 3) (M .16, phase 3 ¼ SD .23; tphases 3 2(49) 4.08, po.001), negative affect carried ¼ À ¼ over from phase 2 to phase 3 (M .54, SD .41; t phase 3 ¼ ¼ phases 3 2(49) .86, po.39). À ¼

2.3. Maternal gaze during modified still-face procedure

Means and standard deviations of maternal gaze variables are shown in Table 2 for phases 1 and 3 of the MSFP. Duration of the mother's gaze toward the infant was significantly Fig. 2 – Duration of positive and negative affect displayed by negatively correlated with the frequency of the mother's gaze infants across modified still-face procedure (MSFP) phases. shifts away from the infant (r .79, gaze toward & gaze shift away ¼À The duration values on the y-axis were adjusted for the total po.001). Infant affect variables were not significantly asso- length of the each respective phase of the MSFP and ciated with either of the maternal gaze variables (rpositive affect represent mean proportion values. Positive affect dropped .09, p .38 and r .09, & gaze toward ¼À ¼ negative affect & gaze toward ¼ and negative affect increased in the still-face segment p .38; r .01, p .91 and r ¼ positive affect & gaze shift away ¼À ¼ negative (phase 2), while a rebound in positive affect and a carry-over .06, p .56), and did not significantly alter affect & gaze shift away ¼ ¼ of negative affect were seen during the recovery period any of the observed main or interaction effects reported nnn (phase 3). po.001. below when entered into the model.

Table 2 – Maternal gaze toward and away from infant during modified still-face procedure (MSFP) and results of mixed- effects regression analysis (N 50). ¼ Maternal gaze during modified still-face Mixed-effects models of change over phase procedure a

Phase 1 Phase 3 Total b Wald χ2 OTResp effect Phase effect OTResp x (df 3) c (95% CI) d (95% CI) d Phase d ¼ (95% CI) Mean SD Mean SD Mean SD

Gaze toward .993 .018 .984 .043 .988 .028 13.52nn .003n .005n .006n À infant (.0002 to .005) ( .009 to .0001) (.001 to .011) À À (duration) Gaze shift away .012 .009 .018 .024 .015 .015 16.14nn .065n .092nn .036 À À from infant ( .126 to .004) (.035 to .150) ( .098 to .026) À À À (frequency)

a Numbers shown are duration and frequency values (M7SD) adjusted for the total length of time for which codable data were available in each respective phase of the MSFP. Untransformed values are reported here for illustrative purposes, while statistical tests were conducted using log-transformed variables. b Total values represent data collapsed over the phases 1 and 3 of the MSFP. c Wald χ2 values are those obtained for the best-fitting mixed-effects models for the respective outcome variables, including a subject-level random intercept and a random coefficient for phase. d Coefficients shown are beta weights (i.e., slopes) for the main and interaction effects of OTResp and phase. n po.05. nn po.01. brain research 1580 (2014) 133–142 137

Fig. 3 – (a) Duration of maternal gaze toward infant and (b) frequency of maternal gaze shift away from infant as a function of maternal oxytocin response (OTResp; mean7SD) and phase of the modified still-face procedure. The y-axes indicate log- transformed estimated marginal means. The statistical significance of the gaze differences between phases 1 and 3 is noted at low (i.e., 1 SD) and average (i.e., mean) levels of maternal oxytocin response.npo.05, nnpo.01 À

2.3.1. Maternal gaze toward infant (duration) are systematically associated with individual variations in The optimal model included a subject-level random intercept mother-to-infant gaze. As hypothesized, maternal peripheral (LR χ2(1) 5.86, p .016) and a random coefficient for phase oxytocin response was positively associated with the dura- ¼ ¼ (LR χ2(2) 38.55, po.0001), providing a significant fit(Table 2). tion of mother-to-infant gaze, while negatively associated ¼ Consistent with our hypothesis, a significant main effect of with the frequency with which maternal gaze was directed OTResp was found, with the duration of mother-to-infant gaze away from infants. Also consistent with our expectation, increasing as OTResp increased across participants. The main these associations were more pronounced under conditions effect of phase was also significant, indicating that the duration of infant distress than non-distress. It is worth noting that of mother-to-infant gaze decreased in phase 3 as compared to mothers with low/average peripheral oxytocin responses phase 1. OTResp and phase interacted significantly. Decomposi- demonstrated a significant decrease in their mother-to- tion of the interaction revealed that a simple effect of phase was infant gaze during periods of infant distress, while such significant at low (i.e., 1 standard deviation below the mean) and change was not observed in mothers with high peripheral average levels of OTResp (β .010, 95% CI .017 oxytocin responses. phase, at low OT ¼À ¼À to .004, z 3.06, p .002 and β .005, 95% The association documented here between maternal oxy- À ¼À ¼ phase, at mean OT ¼À CI .009 to .0001, z 2.00, p .045, respectively), but not tocin and mother-to-infant gaze is consistent with and ¼À À ¼À ¼ significant at high (i.e., 1 standard deviation above the mean) extends previous studies that have reported on the links levels of OTResp (β .001, 95% CI .006 to.007, between maternal oxytocin and other forms of synchronous phase, at high OT ¼ ¼À z .24, p .814) (Fig. 3(a)). This indicated that mothers with high maternal behavior (e.g., affectionate touch; Atzil et al., 2011; ¼ ¼ OTResp displayed gaze duration that was similar during phases Feldman et al., 2010a, 2010b, 2011). In keeping with previous 1and3,whilematernalgazedurationwassignificantly reduced data from animal (Snowdon et al., 2010) and human research for mothers with low/average OTResp during phase 3. (Dawood et al., 1979; de Geest et al., 1985; Feldman et al., 2007; Levine et al., 2007), we found significant individual variations in mothers' peripheral oxytocin responses, spanning from 2.3.2. Maternal gaze shift away from infant (frequency) those that demonstrate a decrease from baseline oxytocin The optimal model was obtained with a random effects level following interactions with infants to those demonstrat- structure that included a subject-level random intercept (LR ing an increase. This variation, likely reflecting naturally χ2(1) 26.75, po.0001) and a random coefficient for phase (LR ¼ occurring differences in mothers' oxytocinergic system func- χ2(2) 8.62, p .013; Table 2). As hypothesized, a significant ¼ ¼ tioning, was associated with two patterns of mother-to- main effect was found for OTResp, with mothers displaying less frequent gaze shifts away from their infants as OTResp infant gaze observed at a micro-behavioral level. Our present fi increased (Fig. 3(b)). The main effect of phase was also nding, in conjunction with the existing literature (Feldman significant, indicating that mothers' gaze shifts from infants et al., 2010a, 2011; Strathearn et al., 2009b, 2012), points to the increased in phase 3 compared to phase 1. OTResp and phase role of oxytocin in regulating the mother's responsive did not interact significantly. engagement with the infant. It is of note that the strength of the observed association between oxytocin and maternal gaze increased during periods of infant distress. From the 3. Discussion inception of attachment theory (Bowlby, 1969/1982), the biological function of mother–infant attachment has been The present results are the first, to our knowledge, to thought to be one of ensuring the infant's access to the document that measures of maternal peripheral oxytocin mother in times of distress (Goldberg et al., 1999; Mikulincer 138 brain research 1580 (2014) 133–142

and Shaver, 2003). Our data underscore the involvement of consisted largely of mothers ofmiddletohighsocioeconomic oxytocin in preparing mothers for such a function. status, which may limit the generalizability of our findings. The significant difference reported here between mothers The present study provides the first evidence for the with high versus low/average oxytocin response is worthy of unique relationship between maternal oxytocin response attention. A corollary to this can be seen in a series of reports and mother-to-infant gaze. We have found results consistent on rodents demonstrating that rat mothers who exhibit low with our understanding that maternal oxytocin may be licking-and-grooming and arched-back nursing (LG–ABN), substantially implicated in the mother's responsive engage- a rodent equivalent of neglectful mothering, showed a reduced ment with her infant, particularly at times when the infant's density of oxytocin receptors in brain regions critical for the need for the mother is greatest. The data presented here may emergence of maternal behavior (i.e., the medial preoptic area, have important implications for intervention in conditions the lateral septum, the paraventricular nucleus of the that challenge optimal mothering. Helping mothers maintain hypothalamus; Champagne et al., 2001; Francis et al., 2000). their engagement during moments of infant distress may be These mothers' low LG–ABN behavior, in turn, was associated an important focus of intervention for mothers whose oxyto- with decreased oxytocin receptor expression in similar brain cin response may be compromised, who suffer from post- regions in the offspring (Champagne et al., 2001, 2003b, 2006) partum depression, or who have struggled with maternal who, like their mothers, subsequently displayed reduced levels substance abuse or trauma. Breastfeeding, though it may not of LG–ABN behavior with their offspring (Champagne et al., lead to long-term changes in baseline oxytocin levels, 2003a; Francis et al., 1999; Lovic et al., 2001). While we have not induces a short-term release of oxytocin in mothers, and examined infant outcomes in the present study, Beebe et al. may hence be beneficial for these mothers (Strathearn et al., (2010)'s micro-analysis of the interaction between mothers 2009a). Finding pharmacological or therapeutic ways to and their 4-month-old infants provides particularly note- enhance maternal oxytocin release may be another impor- worthy information in this regard. Beebe et al. (2010) reported tant focus of future intervention. on a group of mothers whose behaviors strikingly parallel the behaviors of mothers who demonstrated reduced oxytocin responses in the present study. Mothers of infants who went 4. Experimental procedures on to develop profoundly insecure attachment at 12 months were characterized not by their global failure of attunement, 4.1. Participants but rather by a remarkably specific failure to attend to their Participants were 50 first-time mothers aged 19–41 (M 28.07.7) infants' distress. Maternal disengagement from infant distress ¼ is seen as a critical precursor to the formation of disrupted years, recruited through prenatal clinics and community adver- attachment (Allen, 2013). Similarly, the reduced gaze directed tisements as part of a larger study. Of 116 participants initially to infant distress reported here in a subsample of our mothers, recruited during the third trimester of pregnancy, 61 met and those mothers' decreased oxytocin response as a possible eligibility criteria, and 50 completed oxytocin sampling and biological marker for their impoverished gaze and mirroring MSFP at 7 months postpartum. Exclusion criterion included response, may subsequently be linked to less-than-optimal mothers with past or present alcohol or substance abuse, outcomes in their infants, particularly in terms of affect mothers who used cigarettes during pregnancy, and mothers dysregulation (Fonagy et al., 2011). Future research should who were on psychotropic medication at the time of the study. examine the longitudinal links between maternal oxytocin, The institutional review board at Baylor College of Medicine maternal gaze, and infant developmental outcomes. approved the research protocol, and all participants provided Several limitations of the study should be recognized. First, written informed consent. we relied on peripheral measures of oxytocin, using a radio- immunoassay technique. While a degree of concordance has 4.2. Measures and procedure been reported between central and peripheral measures of oxytocin (Carter et al., 2007), peripheral oxytocin levels may 4.2.1. Oxytocin sampling not accurately reflect central oxytocin activity (Amico et al., Mothers attended the study session with their infants 7 1990). This represents a limitation yet to be fully overcome in months postpartum. Mothers were instructed to abstain from human oxytocin research, although we have previously demon- caffeine and tobacco 2–3 h prior to the scheduled visit. During strated a correlation between peripheral oxytocin response and the visit, mother–infant dyads participated in a semi- brain activation in the hypothalamic/pituitary region, where structured mother–infant interaction procedure, which con- oxytocinergic neurons are concentrated (Strathearn et al., sisted of two periods of separation and an intervening period 2009b). Second, we did not track changes in maternal oxytocin of mother–infant interaction. Four serial measurements of or gaze direction as a function of changes in infant affect. We plasma oxytocin were obtained from the mothers during the were therefore unable to examine how maternal oxytocin procedure (Fig. 4). moderated the temporal process by which infant distress modifies and is, in turn, modified by maternal gaze. This would 4.2.1.1. Baseline. Upon an initial separation of mother and be a fruitful area for further investigation. Third, we considered infant, an intravenous cannula was inserted into the oxytocin concentration during the initial mother–infant separa- mother's forearm and a blood sample was taken for estradiol tion as ‘baseline,’ although we acknowledge that this baseline and progesterone levels. Blood was drawn again 20 min later measurement may have been confounded by the stress experi- to determine the first baseline measurement of plasma enced during the separation procedure. Finally, our sample oxytocin (OT 1). brain research 1580 (2014) 133–142 139

Trained raters, who had had no prior contact with study participants and were blind to the study hypotheses, coded the videotaped interactions. Coding of each behavior category was performed in 1-s time intervals during multiple viewings of videotapes (independently for each behavior). Tapes were viewed at normal speed during coding; however, they were often stopped or run in slow motion to confirm codes and accurately determine the beginning and end of episodes of Fig. 4 – Mother–infant interaction procedure and oxytocin maternal and infant behavior. Onset and offset times of response (OTResp) calculation. Measurements of plasma behaviors were entered into a custom-built software, which oxytocin were obtained from mothers at four time points: generated frequency and duration data for each variable of (a) following the first period of mother–infant separation interest. The median inter-rater reliability was r .86 for 18 (OT 1), (b) following two periods of mother–infant interaction s ¼ double-coded interactions (36%). (OT 2 and OT 3), and (c) following the final period of mother– infant separation (OT 4). 4.2.3. Additional mother and infant characteristics Several characteristics of interest for mothers and infants were also examined. The Beck Depression Inventory-II (BDI-II; 4.2.1.2. Post free-play. Following an initial 20-min separation, Beck et al., 1996), the Personality Disorder Questionnaire 4 mother and infant were reunited for a 5-min ‘free-play’ À þ (PDQ-4 ; Hyler et al., 1992), and the Parenting Stress Index period, which included direct physical interaction between þ (PSI; Abidin, 1995) were administered to assess symptoms of mother and infant on the floor with age-appropriate toys. The depression, personality disorders, and parenting stress in second blood sample was then drawn through the previously mothers. We also collected information regarding infants' inserted cannula (OT 2). breastfeeding and daycare status (i.e., if the mother was still breastfeeding or not, and the number of hours per week the 4.2.1.3. Post modified still-face procedure (MSFP). A 6-min infant was cared for by someone other than the mother). MSFP was then conducted. The MSFP is a structured experi- Details on the psychometric properties of all administered mental paradigm whereby the mother interacts with her measures can be found in Shah et al. (2010). infant through three successive phases, including a still- face phase (see Section 4.2.2; Fig. 1). Mother and infant were 4.3. Blood collection and assay able to see each other via a mirror and hear one another during the procedure, but were separated by a dividing screen Blood samples were collected and processed by registered and thus could not physically interact. A third blood sample nurses from the General Clinical Research Center. Estradiol was obtained following the MSFP (OT 3). and oxytocin samples were placed in chilled heparinized tubes and kept on ice. These tubes were centrifuged to 4.2.1.4. Baseline (post). A second period of 20-min separation separate plasma within 2 h after collection, and the plasma occurred between mother and infant, after which the final was rapidly frozen and maintained at 80 1C. Before centri- oxytocin blood sample (OT 4) was obtained. À fuging, blood samples for progesterone were placed in a serum separator tube and allowed to clot at room tempera- 4.2.2. Modified still-face procedure (MSFP) ture. A commercial laboratory determined plasma estradiol The MSFP (Koos and Gergely, 2001) adhered to the standard and serum progesterone concentrations using quantitative still-face procedure (Tronick et al., 1978) with one exception: chemiluminescent immunoassay. Dr. Janet Amico's labora- the mother and infant were seated beside one another (the tory at the University of Pittsburgh received batches of mother in a chair and the infant in a high chair), separated by oxytocin samples that were sent on dry ice by overnight a divider and facing a one-way mirror (Fig. 1(a)). The purpose courier. A sensitive and specific liquid-phase radioimmu- of the divider and one-way mirror was to prevent physical noassay (RIA) of oxytocin in plasma was performed on interaction and touch, but allow mother and infant to see acetone–ether extracted material, using previously published each other reflected in the mirror.1 On the opposing side of and validated methods in which oxytocin antiserum does not the one-way mirror were two cameras, generating a split- cross-react with arginine vasopressin or other oxytocin-like screen recording of the mother and infant. Interactions peptides (Amico et al., 1985). The lower limit of detection for between mother and infant were videotaped during each of this assay is .5 pg/ml and inter- and intra-assay coefficients of the three 2-min phases (Fig. 1(b)): (1) the baseline normal variation are each o10%. Although more labor-intensive than interaction phase (phase 1), (2) the still-face phase, during widely used alternate methods such as the enzyme immu- which the mother was asked to assume a neutral face (phase noassay, RIA on extracted plasma has consistently produced 2), and (3) the recovery phase, in which the mother resumed valid and reliable results (McCullough et al., 2013). free interaction with the infant (phase 3). An intercom was used to communicate the start of each phase to the mother. 4.4. Variables

1The purpose of the modification of the still-face procedure was to test separate hypotheses regarding infant gaze preference 4.4.1. Maternal oxytocin response (OTResp) (self vs. mother), which was not explored in this paper focusing OTResp (Strathearn et al., 2012) was calculated by computing on mother-to-infant gaze. the change in oxytocin concentration between measurement 140 brain research 1580 (2014) 133–142

at baseline (OT 1) and measurement following mother–infant interactions (mean of OT 2 and OT 3; see Fig. 4). OT 2 and OT 3 Funding sources were highly correlated with each other (rs OT2 & OT3 .61, ¼ po.001). OT 4 was omitted from further analyses as a carry- This work was supported by the Eunice Kennedy Shriver over effect was seen from the mother–infant interaction National Institute of Child Health and Human Development [K23 HD43097, R01 HD065819]; General Clinical Research phases; OT 4 was associated with both OT 3 and OT 2 (rs OT3 .41, p .004 and r .43, p .002), but not with Center [MO1 RR00188]; the Baylor Child Health Research & OT4¼ ¼ s OT2 & OT4¼ ¼ baseline (r .15, p .30). Positive OTResp values Center: Pediatrics Mentored Research Program [K12 s OT1 & OT4¼ ¼ indicated a relative increase in oxytocin during mother– HD41648]; and the National Institute on Drug Abuse [R01 infant interactions, while negative OTResp values indicated DA026437]. The content is solely the responsibility of the fi a relative decrease. For four mothers, a single missing authors and does not necessarily represent the of cial views oxytocin value was imputed using linear interpolation. of these institutes or the National Institutes of Health.

4.4.2. Maternal gaze variables We coded maternal eye gaze toward and gaze shifts away Acknowledgments from the infant during the two interactive phases (i.e., phases 1 and 3) of the MSFP. Maternal gaze toward infant was We would like to thank Houming Cai and the late Dr. Janet quantified by the total duration of time, in seconds, that the Amico for assistance in performing the radioimmunoassays mother looked at her infant. Maternal gaze shifts away from of oxytocin. Udita Iyengar and Sheila Martinez provided infant were assessed by the total frequency with which the valuable assistance with data management and analysis. mother's gaze first fixated on the infant (i.e., remained stationary on the infant for a minimum of 1 s), then shifted away from the infant. MSFP phases were recorded for 2 min references each with slight variations in timing due to mothers' com- pliance with procedure instructions and infant behavior. Thus, gaze duration and frequency values were adjusted for Abidin, R., 1995. Parenting Stress Index. Professional Manual, 3rd ed. Psychological Assessments Resources, Lutz, FL. the total length of time in each phase of the MSFP. Allen, J.G., 2013. Mentalizing in the Development and Treatment of Attachment Trauma. Karnac, London, UK. 4.4.3. Infant affect variables Amico, J.A., et al., 1985. A novel oxytocin-like and vasotocin-like Infant facial expressions and vocalizations were coded for peptide in human plasma after administration of estrogen. J. Clin. Endocrinol. Metab. 60, 5–12. positive and negative affect at 1-s intervals throughout the Amico, J.A., Challinor, S.M., Cameron, J.L., 1990. Pattern of MSFP. Total duration of positive and negative affect was oxytocin concentrations in the plasma and cerebrospinal fluid calculated for the three phases of the MSFP, adjusting for of lactating rhesus monkeys (Macaca mulatta): evidence for the total length of time in each respective phase. functionally independent oxytocinergic pathways in primates. J. Clin. Endocrinol. Metab. 71, 1531–1535. Atzil, S., Hendler, T., Feldman, R., 2011. Specifying the 4.5. Statistical analysis neurobiological basis of human attachment: brain, hormones, and behavior in synchronous and intrusive mothers. All variables were inspected for normality via quantile– Neuropsychopharmacology 36, 2603–2615. quantile plots of residuals against fitted values. Logarithmic Beck, A.T., Steer, R.A., Brown, G.K., 1996. Manual for the Beck transformations were performed on maternal gaze variables Depression Inventory II. Psychological Corporation, San to optimize the approximation to normal distribution. Antonio, TX. Beebe, B., et al., 2010. The origins of 12-month attachment: Changes in infant positive and negative affect were examined a microanalysis of 4-month mother–infant interaction. in repeated-measures ANOVAs with MSFP phase as a within- Attach. Hum. Dev. 12, 3–141. subject factor, which were followed by post hoc mean Benarroch, E.E., 2013. Oxytocin and vasopressin: social comparisons. Maternal gaze variables were centered prior to neuropeptides with complex neuromodulatory functions. being submitted to mixed-effects linear regression analysis. Neurology 80, 1521–1528. Model building was carried out as follows. (a) The initial Bowlby, J., 1969/1982. Attachment. Basic Books, New York, NY. model included the fixed main effects of OTResp and MSFP Carter, C.S., et al., 2007. Oxytocin: behavioral associations and potential as a salivary biomarker. Ann. N. Y. Acad. Sci. 1098, phase (phase 1 vs. phase 3). (b) Subject-level random inter- 312–322. cept and slope were added to model systematic inter- Champagne, F., et al., 2001. Naturally occurring variations in individual variability. (c) Interaction terms were added maternal behavior in the rat are associated with differences in sequentially and retained in the model if they improved estrogen-inducible central oxytocin receptors. Proc. Natl. model fit. (d) Infant affect variables were added as covariates Acad. Sci. USA 98, 12736–12741. to examine whether variability in infant affect altered the Champagne, F.A., et al., 2003a. Variations in maternal care in the fl significance of the model fit or parameter estimates. The rat as a mediating in uence for the effects of environment on development. Physiol. Behav. 79, 359–371. model was fitted by maximum likelihood estimation, and Champagne, F.A., et al., 2003b. Natural variations in maternal likelihood-ratio chi-square tests were used to assess the care are associated with estrogen receptor alpha expression relative fit of nested models. All analyses were conducted and estrogen sensitivity in the medial preoptic area. using STATA/SE version 12.1 and SPSS version 21. Endocrinology 144, 4720–4724. brain research 1580 (2014) 133–142 141

Champagne, F.A., et al., 2006. Maternal care associated with Hyler, S.E., et al., 1992. Validity of the Personality Diagnostic methylation of the estrogen receptor-alpha1b promoter and Questionnaire-revised: a replication in an outpatient sample. estrogen receptor-alpha expression in the medial preoptic Compr. Psychiatry 33, 73–77. area of female offspring. Endocrinology 147, 2909–2915. Insel, T.R., 1992. Oxytocin – a neuropeptide for affiliation: evidence Colonnesi, C., et al., 2012. Coordination of gaze, facial expressions from behavioral, receptor autoradiographic, and comparative and vocalizations of early infant communication with mother studies. Psychoneuroendocrinology 17, 3–35. and father. Infant Behav. Dev. 35, 523–532. Keverne, E.B., Kendrick, K.M., 1992. Oxytocin facilitation of Dawood, M.Y., et al., 1979. Oxytocin in maternal circulation and maternal behavior in sheep. Ann. N. Y. Acad. Sci. 652, 83–101. amniotic fluid during pregnancy. J. Clin. Endocrinol. Metab. 49, Kochanska, G., 2001. Emotional development in children with 429–434. different attachment histories: the first three years. Child Dev. de Geest, K., et al., 1985. Plasma oxytocin in human pregnancy 72, 474–490. and parturition. J. Perinat. Med. 13, 3–13. Kochanska, G., Kim, S., 2013. Early attachment organization with Febo, M., Numan, M., Ferris, C.F., 2005. Functional magnetic both parents and future behavior problems: from infancy to resonance imaging shows oxytocin activates brain regions middle childhood. Child Dev. 84, 283–296. associated with mother–pup bonding during suckling. J. Koos, O., Gergely, G., 2001. A contingency-based approach to the Neurosci. 25, 11637–11644. etiology of ‘disorganized’ attachment: the ‘flickering switch’ Feldman, R., 2007. Parent–infant synchrony and the construction hypothesis. Bull. Menn. Clin. 65, 397–410. of shared timing; physiological precursors, developmental Lavelli, M., Fogel, A., 2013. Interdyad differences in early mother– outcomes, and risk conditions. J. Child Psychol. Psychiatry infant face-to-face communication: real-time dynamics and Allied Discipl. 48, 329–354. developmental pathways. Dev. Psychol. http://dx.doi.org/ Feldman, R., et al., 2007. Evidence for a neuroendocrinological 10.1037/a0032268 [epub ahead of print]. foundation of human affiliation: plasma oxytocin levels Leerkes, E.M., Blankson, A.N., O'Brien, M., 2009. Differential across pregnancy and the postpartum period predict mother– effects of maternal sensitivity to infant distress and infant bonding. Psychol. Sci. 18, 965–970. nondistress on socio-emotional functioning. Child Dev. 80, Feldman, R., et al., 2010a. Natural variations in maternal and 762–775. paternal care are associated with systematic changes in Levine, A., et al., 2007. Oxytocin during pregnancy and early oxytocin following parent–infant contact. postpartum: individual patterns and maternal–fetal Psychoneuroendocrinology 35, 1133–1141. attachment. Peptides 28, 1162–1169. Feldman, R., Gordon, I., Zagoory-Sharon, O., 2010b. The cross- Lischke, A., et al., 2012. Intranasal oxytocin enhances emotion generation transmission of oxytocin in humans. Horm. Behav. recognition from dynamic facial expressions and leaves eye- 58, 669–676. gaze unaffected. Psychoneuroendocrinology 37, 475–481. Feldman, R., Gordon, I., Zagoory-Sharon, O., 2011. Maternal and Lovic, V., Gonzalez, A., Fleming, A.S., 2001. Maternally separated paternal plasma, salivary, and urinary oxytocin and parent– rats show deficits in maternal care in adulthood. Dev. infant synchrony: considering stress and affiliation Psychobiol. 39, 19–33. components of human bonding. Dev. Sci. 14, 752–761. Maestripieri, D., et al., 2009. Mother–infant interactions in free- Feldman, R., 2012. Oxytocin and social affiliation in humans. ranging rhesus macaques: relationships between Horm. Behav. 61, 380–391. physiological and behavioral variables. Physiol. Beh. 96, Fonagy, P., Gergely, G., Target, M., 2007. The parent–infant dyad 613–619. and the construction of the subjective self. J. Child Psychol. Mayes, L.C., 2006. Arousal regulation, emotional flexibility, medial Psychiatry Allied Discip. 48, 288–328. amygdala function, and the impact of early experience: Fonagy, P., Luyten, P., Strathearn, L., 2011. Borderline personality comments on the paper of Lewis et al. Ann. N. Y. Acad. Sci. disorder, mentalization, and the neurobiology of attachment. 1094, 178–192. Infant Ment. Health J. 21, 47–69. McCullough, M.E., Churchland, P.S., Mendez, A.J., 2013. Problems Francis, D., et al., 1999. Nongenomic transmission across with measuring peripheral oxytocin: can the data on oxytocin generations of maternal behavior and stress responses in the and human behavior be trusted?. Neurosci. Biobehav. Rev. 37, rat. Science 286, 1155–1158. 1485–1492. Francis, D.D., Champagne, F.C., Meaney, M.J., 2000. Variations in McElwain, N.L., Booth-LaForce, C., 2006. Maternal sensitivity maternal behaviour are associated with differences in to infant distress and nondistress as predictors oxytocin receptor levels in the rat. J. Neuroendocrinol. 12, of infant–mother attachment security. J. Fam. Psychol. 20, 1145–1148. 247–255. Francis, D.D., et al., 2002. Naturally occurring differences in Mesman, J., Van Ijzendoorn, M.H., Bakermans-Kranenburg, M.J., maternal care are associated with the expression of oxytocin 2009. The many faces of the still-face paradigm: a review and and vasopressin (V1a) receptors: gender differences. J. meta-analysis. Dev. Rev. 29, 120–162. Neuroendocrinol. 14, 349–353. Meyer-Lindenberg, A., et al., 2011. Oxytocin and vasopressin in Goldberg, S., Grusec, J.E., Jenkins, J.M., 1999. Confidence in the human brain: social neuropeptides for translational protection: arguments for a narrow definition of attachment. J. medicine. Nat. Rev. Neurosci. 12, 524–538. Family Psychol. 13, 475–483. Mikulincer, M., Shaver, P.R., 2003. The attachment behavioral Gordon, I., et al., 2008. Oxytocin and cortisol in romantically system in adulthood: activation, psychodynamics, and unattached young adults: associations with bonding and interpersonal processes. In: Zanna, M.P. (Ed.), Advances in psychological distress. Psychophysiology 45, 349–352. Experimental and Social Psychology, vol. 35. Academic Press, Gordon, I., et al., 2010. Oxytocin and the development of New York, NY, pp. 52–153. parenting in humans. Biol. Psychiatry 68, 377–382. Mikulincer, M., Shaver, P.R., 2007. Attachment in Adulthood: Haley, D.W., Stansbury, K., 2003. Infant stress and parent Structure, Dynamics, and Change. Guilford Press, New York, NY. responsiveness: regulation of physiology and behavior during Pedersen, C.A., et al., 2006. Maternal behavior deficits in nulliparous still-face and reunion. Child Dev. 74, 1534–1546. oxytocin knockout mice. Genes Brain Behav. 5, 274–281. Hurlemann, R., et al., 2010. Oxytocin enhances amygdala- Perry, A., et al., 2013. Face or body? Oxytocin improves perception dependent, socially reinforced learning and emotional of emotions from facial expressions in incongruent emotional empathy in humans. J. Neurosci. 30, 4999–5007. body context. Psychoneuroendocrinology 38, 2820–2825. 142 brain research 1580 (2014) 133–142

Rodrigues, S.M., et al., 2009. Oxytocin receptor genetic variation Strathearn, L., et al., 2009a. Does breastfeeding protect against relates to empathy and stress reactivity in humans. Proc. Natl. substantiated child abuse and neglect? A 15-year cohort Acad. Sci. USA 106, 21437–21441. study. Pediatrics 123, 483–493. Ross, H.E., et al., 2009. Variation in oxytocin receptor density in Strathearn, L., et al., 2009b. Adult attachment predicts maternal the nucleus accumbens has differential effects on affiliative brain and oxytocin response to infant cues. behaviors in monogamous and polygamous voles. J. Neurosci. Neuropsychopharmacology 34, 2655–2666. 29, 1312–1318. Strathearn, L., 2011. Maternal neglect: oxytocin, dopamine and the Schneiderman, I., et al., 2012. Oxytocin during the initial stages of neurobiology of attachment. J. Neuroendocrinol. 23, 1054–1065. romantic attachment: relations to couples' interactive Strathearn, L., et al., 2012. Maternal oxytocin response during reciprocity. Psychoneuroendocrinology 37, 1277–1285. mother–infant interaction: associations with adult Schore, A.N., 2001. Effects of a secure attachment relationship on temperament. Horm. Behav. 61, 429–435. right brain development, affect regulation, and infant mental Tronick, E., et al., 1978. The infant's response to entrapment health. Infant Ment. Health J. 22, 7–66. between contradictory messages in face-to-face interaction. J. Shah, P.E., Fonagy, P., Strathearn, L., 2010. Is attachment Am. Acad. Child Psychiatry 17, 1–13. transmitted across generations? The plot thickens. Clin. Child Tronick, E., Als, H., Brazelton, T.B., 1980. Monadic phases: a Psychol. Psychiatry 15, 329–345. structural descriptive analysis of infant–mother face to face Slee, P.T., 1984. The nature of mother–infant gaze patterns during interaction. Merrill-Palmer Q. 26, 3–24. interaction as a function of emotional expression. J. Am. Acad. Van IJzendoorn, M.H., Bakermans-Kranenburg, M.J., 2012. A sniff Child Psychiatry 23, 385–391. of trust: meta-analysis of the effects of intranasal oxytocin Snowdon, C.T., et al., 2010. Variation in oxytocin is related to administration on face recognition, trust to in-group, and variation in affiliative behavior in monogamous, pairbonded trust to out-group. Psychoneuroendocrinology 37, 438–443. tamarins. Horm. Beh. 58, 614–618. Weinfield, N.S., Whaley, G.J., Egeland, B., 2004. Continuity, Sroufe, L.A., et al., 2005. The Development of the Person. Guilford discontinuity, and coherence in attachment from infancy to Press, New York, NY. late adolescence: sequelae of organization and Stern, D.N., 1974. Mother and infant at play: the dyadic disorganization. Attach. Hum. Dev. 6, 73–97. interaction involving facial, vocal and gaze behaviors. In: Williams, G.L., et al., 2001. Physiological regulation of maternal Lewis, M., Rosenblum, L.A. (Eds.), The Effect of the Infant on behavior in heifers: roles of genital stimulation, intracerebral its Caregiver. John Wiley & Sons, New York, NY, pp. 187–215. oxytocin release, and ovarian steroids. Biol. Reprod. 65, 295–300. brain research 1580 (2014) 219–232

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Oxytocin and postpartum depression: Delivering on what's known and what's not

Sohye Kima,b,c, Timothy A. Soekena, Sara J. Cromera, Sheila R. Martineza,b, Leah R. Hardya,b, Lane Strathearna,b,c,d,n aBaylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA bAttachment and Neurodevelopment Laboratory, Children's Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, 1100 Bates Street, Suite 4004, Houston, TX 77030, USA cMenninger Department of Psychiatry and Behavioral Sciences, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030, USA dThe Meyer Center for Developmental Pediatrics, Texas Children's Hospital, 8080 N. Stadium Drive, Houston, TX 77054, USA article info abstract

Article history: The role of oxytocin in the treatment of postpartum depression has been a topic of growing Accepted 7 November 2013 interest. This subject carries important implications, given that postpartum depression can Available online 14 November 2013 have detrimental effects on both the mother and her infant, with lifelong consequences for Keywords: infant socioemotional and cognitive development. In recent years, oxytocin has received Postpartum depression attention for its potential role in many neuropsychiatric conditions beyond its well- Oxytocin described functions in childbirth and lactation. In the present review, we present available Treatment data on the clinical characteristics and neuroendocrine foundations of postpartum Maternal caregiving depression. We outline current treatment modalities and their limitations, and proceed Infant to evaluate the potential role of oxytocin in the treatment of postpartum depression. The aim of the present review is twofold: (a) to bring together evidence from animal and human research concerning the role of oxytocin in postpartum depression, and (b) to highlight areas that deserve further research in order to bring a fuller understanding of oxytocin's therapeutic potential. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2013 Elsevier B.V. All rights reserved.

1. Introduction following childbirth, posing numerous challenges to the mother- infant relationship and the infant's subsequent development Postpartum depression (PPD) is a debilitating disorder affecting (Murray and Cooper, 1997). Despite the increased attention at least one in seven American women annually (Gaynes given to PPD over the past several decades, PPD is currently et al., 2005). PPD impairs the mother's capacity for adaptation conceptualized and treated in much the same manner as

nCorresponding author at: Baylor College of Medicine/Texas Children's Hospital, Department of Pediatrics, Attachment and Neurodevelopment Laboratory, Children's Nutrition Research Center, 1100 Bates St, Suite 4004-B, Houston, TX 77030, USA. Fax: 1 713 798 7009. þ E-mail address: [email protected] (L. Strathearn).

0006-8993/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.brainres.2013.11.009 220 brain research 1580 (2014) 219–232

non-postpartum depression, often leading to less than optimal criteria used, particularly the time criterion. However, find- treatment outcomes. In a separate but related line of research, a ings from meta-analytic and systematic reviews converge to substantial interest has centered around oxytocin (OT), a neuro- point to a more precise estimate of 10 to 15% (Gaynes et al., peptide hormone critically implicated in the transition to and 2005; O'Hara and Swain, 1996), translating to approximately adjustment during early motherhood. The past two decades 600,000 women in the United States annually. This is distin- have witnessed a surge of clinical trials evaluating OT's ther- guished from the postpartum blues, a mild and transient apeutic potential in a wide range of psychiatric disorders, and mood disturbance following childbirth, commonly experi- increasing attention is now directed to PPD as another clinical enced by up to 80% of postpartum women (Beck, 2006; syndrome for which OT may be of therapeutic benefit. In the Buttner et al., 2012). The risk of PPD increases with a history present review, we draw upon available data from animal and of prenatal depression, prenatal anxiety, or PPD (Beck, 2001; human research to critically evaluate the potential role of OT in Robertson et al., 2004; Wisner and Wheeler, 1994). Stressors treating PPD. Our goal is to bring together relevant evidence that during pregnancy and the early postpartum, including peri- may elucidate the potential of OT as a therapeutic agent for PPD, natal complications, preterm birth, or infant health problems while highlighting areas where further research is necessary. (Blom et al., 2010; Robertson et al., 2004; Sit and Wisner, 2009), also serve to increase the risk of PPD, as do poverty, low social support, and adolescent motherhood (Beck, 2001; O'Hara and 2. Postpartum depression Swain, 1996; Robertson et al., 2004; Troutman and Cutrona, 1990; Wang et al., 2011). PPD lasts for more than 7 months in 2.1. Definition and diagnosis over half of affected women (Sit and Wisner, 2009). While studies have demonstrated a high heritability of PPD is defined as the presence of a major depressive episode depressive disorders (Sullivan et al., 2000), evidence is less following childbirth, although there remains controversy conclusive concerning PPD (Corwin et al., 2010). The only regarding the time criterion pertaining to its onset. The published twin study of PPD is one by Treloar et al. (1999), Diagnostic and Statistical Manual of Mental Disorders (DSM)- who demonstrated that genetic factors accounted for 25% of IV utilized the specifier “with postpartum onset” to limit the variance in the onset of PPD in 838 Australian female twin diagnosis of PPD to depressive episodes manifesting within pairs. Three family studies exist to date and suggest that the 4 weeks post delivery (American Psychiatric Association, 2000), rate of PPD increases in female siblings of women with while the newly released DSM-5 uses the specifier “with unipolar (Forty et al., 2006; Murphy-Eberenz et al., 2006) or peripartum onset” to encompass depressive episodes present bipolar depression (Payne et al., 2008). Although informative, during pregnancy (American Psychiatric Association, 2013). In these few studies have been criticized by some for methodo- contrast to the DSM, the International Classification of Dis- logical shortcomings (e.g., failure to distinguish between PPD eases (ICD)-10 classifies depression “as associated with the and postpartum blues or to control for other psychiatric puerperium” if the onset is within 6 weeks postpartum (World comorbidity; Corwin et al., 2010), particularly in light of the Health Organization, 1992). In the face of a lack of consensus, lack of association shown by other groups between a two large-scale epidemiological studies have demonstrated woman's familial history of depression and her development that women's risk for psychiatric illness increased from child- of PPD (Bloch et al., 2005; Dennis et al., 2004). Studies of birth to approximately 3 months postpartum; the risk genetic markers have also been underway and have high- increased up to 5 months postpartum specifically for depres- lighted the role of the polymorphisms of three candidate sion (Kendell et al., 1987; Munk-Olsen et al., 2006). In con- genes, the transporter, monoamine oxidase A, and sideration of the epidemiological findings and challenges in catechol-O-methyltransferase genes (Doornbos et al., 2009; practical clinical applications, some experts have recom- Sanjuan et al., 2008), although complex genetic and epige- mended that the time criterion be extended to 3 to 6 months netic interactions remain to be explored. postpartum (Elliott, 2000; Wisner et al., 2010). As with non-postpartum depression, depressive symp- 2.3. Neuroendocrine considerations toms must be present for more than 2 weeks to warrant the diagnosis of PPD. Common symptoms include depressed Elucidating the neuroendocrinology of PPD has been a challenge mood, loss of interest and energy, changes in sleep or eating in the field due to normative and adaptive neuroendocrine patterns, diminished ability to think or concentrate, feelings changes that take place during pregnancy and postpartum. of worthlessness, and recurrent suicidal ideations. While not Gonadal steroid hormones have received attention, as levels of currently a part of diagnostic criteria, anxiety is considered a estradiol and progesterone drop drastically following parturi- prominent feature of PPD, present in approximately half of tion, often coinciding with the onset of postpartum blues or PPD women diagnosed with PPD (Ross et al., 2003). In severe cases, symptoms. A noteworthy study by Bloch et al. (2000) found that PPD can be accompanied by psychotic features which may the simulation of gonadal withdrawal precipitated depressive include delusions or command hallucinations to harm the symptoms in euthymic women withahistoryofPPD,support- infant (American Psychiatric Association, 2013). ing the potential contribution of hypogonadism to the onset of PPD. However, the relationship between hypogonadism and 2.2. Prevalence and risk factors PPD has been disputed by many others (Abou-Saleh et al., 1998; Harris et al., 1996; Klier et al., 2007; Zonana and Gorman, 2005). Prevalence estimates of PPD range widely from 5 to 25% Currently, there is no consistent evidence that a decrease in (Gavin et al., 2005), primarily due to the variability in the absolute concentrations of gonadal hormones triggers PPD, brain research 1580 (2014) 219–232 221

although available data suggest that PPD may manifest in emotional, behavioral, and medical problems. Cognitively, women who are vulnerable to fluctuations in gonadal hormone they are likely to have lower IQ scores, attention problems, levels (Workman et al., 2012). and special educational needs (Hay et al., 2001). Emotionally, Dysregulation of the hypothalamic–pituitary–adrenal (HPA) they are susceptible to various forms of psychopathology axis has also garnered interest for its potential role in the including mood disorders, anxiety disorders, and substance etiology of PPD. The HPA system undergoes numerous changes use disorders (Apter-Levy et al., 2013; Murray et al., 2011; during pregnancy and postpartum (Lightman et al., 2001; Tu Schwartz et al., 1990). Behavioral problems are also prevalent, et al., 2006). Adreno-corticotropic hormone levels increase during at times warranting the diagnoses of conduct disorder or pregnancy, and cortisol reaches its peak at the end of pregnancy oppositional defiant disorder (Alpern and Lyons-Ruth, 1993; as the placental corticotropin-releasing hormone (CRH) levels Dawson et al., 2003). They are also often high utilizers of rise, before dropping rapidly at parturition (Kammerer et al., pediatric emergency services, while frequently missing out- 2006; Yim et al., 2009). There have been reports that women with patient pediatric visits (Flynn et al., 2004). Many of these PPD demonstrate more extreme changes in the activity of the adverse developmental outcomes have been associated with HPA axis during pregnancy and postpartum (Jolley et al., 2007; impaired maternal caregiving behavior in depression, even Taylor et al., 2009), though directionality has been inconsistent, after controlling for the effects of demographic variables with increased (Lommatzsch et al., 2006; Okano and Nomura, (Azak and Raeder, 2013; NICHD Early Child Care Research 1992)ordecreasedcortisollevelsbeingdocumented(Groer and Network, 1999). Morgan, 2007; Jolley et al., 2007). Indeed, some have proposed A large number of studies have highlighted that mothers that different subtypes of PPD may exist, underpinned by distinct with PPD are slow to read, decipher, and respond to their genetic predispositions and differential regulation patterns infants' signals (see reviews, Field, 2010; Tronick and Reck, (i.e., hypo- vs. hyper-regulation) of the HPA axis (Kammerer 2009). While some depressed mothers are withdrawn, pas- et al., 2006). sive, and under-stimulating, others are intrusive, hostile, and OT has received less interest than gonadal or stress hor- over-stimulating (Lovejoy et al., 2000; Malphurs et al., 1996; mones as a potential etiologic factor in PPD, although it has Weikum et al., 2013). This provides infants with fewer and attracted attention for its involvement in breastfeeding difficul- shorter moments of reciprocal engagement, joint attention, ties often present in PPD (Stuebe et al., 2012). It is well known and shared affect (Feldman, 2007; Weinberg and Tronick, that OT is critically implicated in milk letdown (Pang and 1998). Infants are also given fewer opportunities to experi- Hartmann, 2007). The documented association between breast- ence repair following moments of disrupted engagement feeding difficulties and PPD (Dennis and McQueen, 2009; (Jameson et al., 1997). Such disruptions in the early mother– Taveras et al., 2003; Watkins et al., 2011)isworthyofattention infant relationship may serve as a precursor to insecure (Skalkidou et al., 2010). To date, only one study (Stuebe et al., infant attachment (Mills-Koonce et al., 2008; Stern, 1995), 2013)hasexaminedOTaspartofthelinkbetweenlactation which has longitudinally been linked to numerous adverse failure and PPD. In a group of mothers intending to breastfeed, developmental outcomes (Sroufe et al., 2005), including the the authors found that OT levels were inversely correlated with cognitive, emotional, and behavioral outcomes described depressive symptoms in both the third trimester and at 8 weeks above. One recent study (Laurent and Ablow, 2013) documen- postpartum, corroborating and extending earlier findings by ted that maternal brain responses were altered in women Skrundz et al. (2011).TheauthorsalsodocumentedthatOT with PPD. While viewing images of their own infants, release was reduced in depressed mothers during breastfeeding mothers with PPD displayed blunted activity in brain regions compared to non-depressed mothers, although no difference known to be central for emotional responsiveness, empathy, was found between the two groups in breastfeeding duration or and reward (e.g., anterior cingulate, orbitofrontal cortex, intensity. Stress-attenuating effects of breastfeeding deserve insula, and striatum). consideration here (Heinrichs et al., 2001; Stuebe et al., 2012). Frequently, the impaired mother–infant relationship is Compared to their non-breastfeeding counterparts, breastfeed- further compromised by breastfeeding difficulties often experi- ing women demonstrated attenuated HPA response to stressors enced in women with PPD. While breastfeeding provides one of (Altemus et al., 1995). Furthermore, cortisol levels decreased in the earliest opportunities for the mother to establish an breastfeeding women during lactation (Amico et al., 1994), while intimate bond with her infant, many depressed mothers report mood scores improved following lactation (Heinrichs et al., dissatisfaction with breastfeeding (Dennis and McQueen, 2007) 2001). Animal models (Neumann, 2003)havesuggestedthat and discontinue breastfeeding between 4 to 16 weeks postpar- disruptions of the OT system may be implicated in the observed tum (McLearn et al., 2006; Paulson et al., 2006). relations between breastfeeding, stress regulation, and mood, which is of particular relevance to PPD given that all three components become disrupted in many affected women 4. Current treatment of postpartum depression (Heinrichs et al., 2001; Stuebe et al., 2012). 4.1. Pharmacotherapy

3. Maternal caregiving in postpartum depression 4.1.1. Antidepressant medication PPD is currently treated in much the same manner as non- PPD is well known to have deleterious effects on the devel- postpartum depression, with selective serotonin reuptake opment of the offspring. Reports show that children of inhibitors (SSRIs) as the first-line of pharmacotherapy. As in depressed mothers are at risk for a wide range of cognitive, non-postpartum depression (Cipriani et al., 2007), SSRIs 222 brain research 1580 (2014) 219–232

are efficacious in targeting depressive symptoms in PPD. remission within 2 weeks of treatment with sublingual estradiol. A review of available randomized controlled clinical trials Notably, decreases in women's depression scores were inversely show that SSRIs improved mood in 43 to 88% of women with correlated with increases in their serum estrogen levels, and PPD (Appleby et al., 1997; Bloch et al., 2012; Misri et al., 2004; available data suggest that estradiol therapy was well tolerated Sharp et al., 2010; Wisner et al., 2006; Yonkers et al., 2008), with low dropout rates. However, estradiol therapy may interfere which are similar to rates reported for non-postpartum with breastfeeding, a consideration important in the treatment depression (Kirsch et al., 2008); 37 to 65% of treated women of PPD (Fitelson et al., 2010; Moses-Kolko et al., 2009). Although achieved remission of depressive symptoms (De Crescenzo early data suggested that synthetic may be et al., 2013). Despite the demonstrated efficacy, factors unique therapeutic for PPD, this view has since been challenged and to pregnancy and the postpartum period complicate the the use of progesterones is not recommended (Dennis et al., antidepressant treatment of PPD (Ellfolk and Malm, 2010; 2008). Currently, no data exist on the effects of hormonal therapy Yonkers et al., 2009). Many healthcare providers are reluctant on maternal behavior. to engage patients in pharmacotherapy due to concerns about fetal or infant exposure to , and three-quarters 4.2. Psychological interventions of women diagnosed with PPD are indeed left untreated (Bennett et al., 2004). Pregnant or postpartum women similarly Psychological interventions that have been empirically stu- show low acceptability of antidepressant medications (Chabrol died for the treatment of PPD include interpersonal psy- et al., 2004), with available SSRI trials reflecting high drop-out chotherapy (IPT; Grote et al., 2009; Mulcahy et al., 2010), rates (Appleby et al., 1997; Wisner et al., 2006; Yonkers et al., cognitive behavioral therapy (CBT; Chabrol et al., 2002; 2008). Medications may also be prescribed and used at sub- Wiklund et al., 2010), psychodynamic therapy (Bloch et al., therapeutic doses (Bennett et al., 2004; Epperson et al., 2003), 2012; Cooper et al., 2003), and non-directive counseling which may be a particular concern since women with PPD (Milgrom et al., 2005; Murray et al., 2003). Several meta- require higher doses of antidepressant agents for a longer analyses exist to date and suggest that psychological inter- duration to experience relief of symptoms (Dawes and ventions are efficacious in reducing depressive symptoms Chowienczyk, 2001; Hendrick et al., 2000). (Cuijpers et al., 2008; Dennis and Hodnett, 2007; Lumley et al., Data on the effects of antidepressant treatment on 2004; Sockol et al., 2011) at rates similar to those of anti- maternal functioning are limited. Only one group depressant medications (De Crescenzo et al., 2013; Sockol (Logsdon et al., 2009, 2011)hasprovidedrelevantdata, et al., 2011). No conclusive data exist to suggest the super- demonstrating that antidepressants were effective in iority of one psychotherapy modality to another in the enhancing maternal role-gratification throughout the first treatment of PPD (Fitelson et al., 2010), although there is postpartum year, but not maternal self-efficacy, overall some evidence that IPT, which directly addresses interperso- maternal role functioning, or the quality of the mother– nal problems (e.g., role transitions, relational conflicts), may infant relationship. The literature on fetal or infant expo- be more efficacious than CBT, which targets maladaptive sure to SSRIs is large and growing. As per a recently depressogenic cognitions (Bledsoe and Grote, 2006; Sockol published report from the American Psychiatric Associa- et al., 2011). tion (APA) and the American College of Obstetricians and However, as with pharmacotherapy, a mere decrease in Gynecologists (ACOG; Yonkers et al., 2009), the accumu- depressive symptoms is often not enough to enhance mater- lated data are in support of small but significant associa- nal functions (Forman et al., 2007; Murray et al., 2003). tions between SSRI use during pregnancy and preterm Psychological interventions without an explicit focus on the delivery (i.e., o37 weeks gestation) as well as small-for- mother–infant relationship were effective in reducing mater- gestational-age birth weight (i.e., o10% of age-adjusted nal parenting stress, without benefiting maternal or infant birth weight). Antidepressant use was generally not asso- behavior (Forman et al., 2007; O'Hara et al., 2000). Interven- ciated with major congenital malformations in the infant, tions that actively incorporated the mother–infant relation- although use during the third trimester was linked to ship as a focus generally yielded some improvement in this transient poor neonatal outcomes (e.g., irritability, domain (Clark et al., 2003; O'Hara and McCabe, 2013; Poobalan jitteriness). et al., 2007). However, therapeutic gains were not sustained on longitudinal follow-ups and did not generalize to infants' 4.1.2. Hormonal therapy cognitive or behavioral outcomes, with the exception of one Estradiol therapy is a novel treatment that targets the flux of study in which extensive and prolonged therapy was imple- gonadal hormones that may render a subset of women vulner- mented (Cicchetti et al., 2000). able to depression in the postpartum period (Bloch et al., 2000; Current APA and AGOC guidelines recommend psy- Moses-Kolko et al., 2009). Growing data suggest that estradiol chotherapy as the first-line of treatment for mild to moderate therapy has promising antidepressant properties, with reported depression, although antidepressant medications are recom- response and remission rates exceeding those of SSRIs. Within mended in the presence of moderate to severe depressive the first month of a double-blinded, placebo-controlled trial symptoms, particularly in women with a history of recurrent (Gregoire et al., 1996), women treated with estrogen patches depression (Kim et al., 2010; Yonkers et al., 2009). The showed greater and more rapid improvements in their depres- acceptability of psychotherapy is reported to be high in sive symptoms compared to those treated with placebo patches. postpartum women (Chabrol et al., 2004) and psychotherapy Similar improvements were shown in another study (Ahokas is often preferred to antidepressants in this population et al., 2001), in which 83% of severely depressed women reached (Pearlstein et al., 2006; Turner et al., 2008). brain research 1580 (2014) 219–232 223

modulating the activity of the HPA axis (Engelmann et al., 5. Oxytocin: A novel therapeutic for postpartum 2004; Neumann et al., 2000). OT knockout mice showed depression? increased stress-induced fos expression in the medial amyg- dala and BNST, heightened CRH mRNA expression in the As the preceding review suggests, currently available treat- PVN, and elevated corticosterone release following exposure ments of PPD are promising in reducing depressive symp- to stressors (Amico et al., 2008; Nomura et al., 2003). Further- toms, but are less effective in improving the mother–infant more, intracerebroventricular infusions of OT decreased the relationship. OT has emerged as a potentially viable treat- molecular and neuroendocrine responses of the HPA axis and ment option in this context, given its role in regulating the attenuated anxiety-like behaviors in female rats, while onset and maintenance of maternal behavior, along with its administration of the OT produced oppo- antidepressant and anxiolytic properties. site results (Slattery and Neumann, 2010a; Windle et al., OT is a nonapeptide synthesized in magnocellular neu- 2004). rons of the paraventricular (PVN) and supraoptic nuclei of the While mechanisms of the antidepressant and anxiolytic hypothalamus and released into the bloodstream from the effects of OT remain to be further elucidated, available neurohypophysis (Gimpl and Fahrenholz, 2001; Insel, 2010). evidence suggests that the effects may be produced in part OT receptors are located throughout the brain including by the interactions between the , corticotropin- regions known to be critical for the expression of maternal releasing factor (CRF), and OT systems (Neumann and behaviors, such as the ventromedial nucleus of the hypotha- Landgraf, 2012). OT release is understood to activate OT lamus, central nucleus of the amygdala, medial preoptic area receptors in serotonergic neurons of the raphe nuclei to yield (MPOA), bed nucleus of the stria terminalis (BNST), and antidepressant and anxiolytic effects (Yoshida et al., 2009), ventral tegmental area (VTA). OT is also secreted in small whereas stimulation of serotonin release activates CRF and amounts by numerous peripheral tissues such as the uterus, OT neurons in the hypothalamus (Javed et al., 1999), a placenta, corpus luteum, testis, and heart (Gimpl and mechanism that may underlie antidepressant properties of Fahrenholz, 2001). For many decades, OT was well recognized the SSRIs (Emiliano et al., 2007). for its peripheral actions, including uterine contraction in parturition and milk ejection during lactation (Insel, 2010; 5.1.2. Human studies Ross and Young, 2009), though it is now regarded a key OT-related dysfunctions have been examined in depressed neuroregulator implicated in social and stress-related disor- patients, although results remain inconclusive. Some groups ders on one hand and maternal behavior on the other (Bartz have shown that peripheral OT concentrations were lower in and Hollander, 2006; Neumann and Landgraf, 2012). To assist depressed patients, particularly female patients (Ozsoy et al., in evaluating the potential of OT as a therapeutic agent for 2009), compared to controls (Frasch et al., 1995; Zetzsche et al., PPD, we examine available animal and human research on 1996). Another report showed that the severity of patients' the role of OT in (a) anxiety- and depressive-like behavior and depression and anxiety symptoms was inversely correlated in (b) maternal functioning. with their plasma OT levels (Scantamburlo et al., 2007). How- ever, other studies of plasma OT (Cyranowski et al., 2008; van 5.1. Antidepressant and anxiolytic effects of oxytocin Londen et al., 1997)andafewavailablestudiesofcerebrospinal fluid (CSF) OT (Demitrack and Gold, 1988; Pitts et al., 1995)failed 5.1.1. Animal studies to document reduced OT levels in depressed patients. Notably, An antidepressant effect of OT was first reported by Arletti these studies found a greater variability of OT concentrations in and Bertolini (1987), who demonstrated that acute and the patient group (Cyranowski et al., 2008; van Londen et al., repeated intra-peritoneal injections of OT in mice reduced 1997); in one study, a trend toward reduced OT was found only the immobility time in the forced swim test, a commonly in a subgroup of patients (Pitts et al., 1995). used index of depressive behavior in animals. Similarly, Endogenous OT released during breastfeeding (Chiodera subcutaneous OT infusions in rats reduced the escape fail- et al., 1991) has been understood to underlie the attenuated ures in the learned helplessness test, another widely used HPA responsiveness and reduced anxiety behavior shown in animal model of depression (Nowakowska et al., 2002). lactating women (Heinrichs et al., 2001). However, exogenous Recently, these results were replicated following intracerebral OT administrations in humans have produced equivocal find- administrations of OT (Ring et al., 2010), suggesting that ings. Intranasal OT was reported to decrease stress respon- central or peripheral administrations of OT may have anti- siveness and anxiety in healthy men (Heinrichs et al., 2003), depressant properties (Slattery and Neumann, 2010b). In reduce levels of salivary cortisol during couple conflict in rats, OT has also been shown to improve other features of heterosexual couples (Ditzen et al., 2009), and attenuate fear- depression (Neumann and Landgraf, 2012), including anhe- related amygdala reactivity in healthy males and patients with donia (Liberzon et al., 1997), sexual dysfunction (Melis et al., social anxiety (Kirsch et al., 2005; Labuschagne et al., 2010). 2007), and sleep disturbance (Lancel et al., 2003). However, these studies failed to document direct effect of Anxiolytic and stress-attenuating effects of OT have also intranasal OT on mood (Kirsch et al., 2005; Labuschagne et al., been well documented. Anxiogenic stimuli are understood to 2010). Furthermore, while the majority of the studies were increase central OT release in the PVN of the hypothalamus conducted in men, intranasal OT was shown to enhance fear- (Nishioka et al., 1998), central nucleus of the amygdala related amygdala reactivity in healthy women (Domes et al., (Ebner et al., 2005), and lateral septum (Ebner et al., 2000), 2010; but see Rupp et al., 2012 for results in nulliparous women which subsequently functions to dampen stress response by only), while producing null effects in healthy postpartum 224 brain research 1580 (2014) 219–232

women (Rupp et al., 2012). Notably, the only available study on thought to mediate the rewarding and reinforcing properties PPD demonstrated that intranasal OT worsened self-reported of the mother–infant interaction. mood ratings in this group of women (Mah et al., 2013). A study observing the use of OT to aid the progress of labor 5.2.2. Human studies further showed that OT administration during labor did not Over the past decade, many important advances have been reduce the incidence of PPD in first-time mothers (Hinshaw extended from animal models to humans, elucidating the et al., 2008). central role of OT in human mothering. Peripheral OT levels in mothers have been consistently associated with naturally 5.2. Oxytocin and maternal behavior occurring variations in maternal behavior, with high OT levels during pregnancy and postpartum predicting enhanced 5.2.1. Animal studies maternal behavior (Atzil et al., 2011; Feldman et al., 2007; A large body of research supports the role of OT in the onset Gordon et al., 2010a, 2010b). Following the work of Meaney and maintenance of maternal behavior across species. Intra- (2001) and Champagne (2008), research in this area has cerebroventricular injections of OT induced a full range of underscored the interindividual variability of OT-related maternal behavior in female virgin rats (Pedersen and Prange, functions in mothers. Interactions with infants stimulate 1979; Pedersen et al., 1982), whereas infusions of OT antago- OT release in mothers, but only in a subgroup of mothers nist inhibited the emergence of maternal behavior in parturi- who demonstrate secure attachment (Strathearn et al., 2009), ent rat dams (Pedersen et al., 1994; van Leengoed et al., 1987). display sensitivity to emotions and physical sensations Similar results were found in mice (McCarthy, 1990) and (Strathearn et al., 2012), or exhibit synchronous and affec- sheep (Kendrick et al., 1987; Keverne and Kendrick, 1992). tionate forms of mothering (Feldman et al., 2010a; Kim et al., These findings are in line with the report that lesions of the 2014). Maternal OT increase during mother–infant inter- PVN, a main site of OT production in the brain, disrupted the action has further been correlated with the concurrently onset of maternal behavior in rats (Insel and Harbaugh, 1989). measured OT increase in the infant, supporting an interge- Impaired maternal behavior was similarly found in female nerational link between the OT functions of mother and OT knockout mice (Ragnauth et al., 2005; Takayanagi et al., infant in humans (Feldman et al., 2010b). Evidence continues 2005) and postpartum mutant female mice with reduced OT to grow supporting the understanding that women's OT neurons in the PVN (Li et al., 1999). Other prominent sites of functions may be modified by their early caregiving experi- OT receptors have also been examined in relation to the ences. Inverse associations have been reported between expression of maternal behavior. OT receptor density in the women's levels of CSF OT and the severity and duration of central nucleus of the amygdala and BNST was shown to be abuse and neglect to which they were exposed in childhood correlated with the quality of maternal care in rats (Francis (Heim et al., 2009). Studies have further documented low et al., 2000); OT receptor binding density in the nucleus plasma OT levels in individuals who reported receiving low accumbens was similarly associated with the amount of time levels of parental care (Feldman et al., 2011; Gordon et al., prairie voles spent crouching over pups (Olazabal and Young, 2008). 2006). Only a handful of studies have examined the role of OT in Another important line of research in this area concerns maternal brain responses. Strathearn et al. (2009) found that the modification of the OT system by early caregiving experi- mesocorticolimbic dopaminergic reward regions (i.e., ventral ences (Champagne, 2008; Meaney, 2001). A series of experi- striatum, medial prefrontal cortex) as well as the hypothalamic mental and cross-fostering studies demonstrated that female OT regions were activated when securely attached mothers rats reared by low licking-and-grooming and arched-back viewed images of their own infants. Similar results were reported nursing (LG-ABN) mothers showed a reduced density of OT by Atzil et al. (2011),whodemonstratedthatmotherswho receptors in brain regions critical for the expression of displayed synchronous forms of mothering showed activation maternal behavior, including the MPOA, PVN, and lateral of the ventral striatum, a key reward region, while viewing video septum (Champagne et al., 2001, 2003b, 2006). Just like their clips of their own infants. Notably, activations in these brain mothers, the female rat pups were subsequently seen to regions were correlated with the peripheral measures of OT in display low levels of LG-ABN behavior with their offspring these mothers (Atzil et al., 2011; Strathearn et al., 2009). Brain when they reached the postpartum period (Champagne et al., responses of breastfeeding mothers have also attracted atten- 2003a; Francis et al., 1999). Similarly, non-maternal rearing in tion, given the role of breastfeeding in endogenous OT produc- rhesus monkeys was associated with reduced CSF OT levels tion (Chiodera et al., 1991). The only pertinent study to date is across the first three years of life (Winslow et al., 2003). that by Kim et al. (2011),whofoundthatbreastfeedingmothers Studies on OT-related maternal functions are continuing showed greater activation of the brain regions critical for the to expand. The current understanding is that OT circuits expression of maternal behavior,includingthestriatalreward interact closely with dopaminergic circuits to regulate the region, in response to their own infants' cries. expression of maternal behavior (Shahrokh et al., 2010; Studies have recently begun to examine the role of Strathearn, 2011). OT neurons in the PVN and MPOA of exogenous OT administrations in maternal brain responses. the hypothalamus project to the VTA and nucleus accum- Intranasal administrations of OT were shown to increase bens (Numan and Smith, 1984; Ross et al., 2009a), and the the incentive salience of an unknown infant's laughter in connections and signals between these regions increase with a group of women, as evidenced by the enhanced connectiv- the quality of maternal behavior (Champagne et al., 2004; ity observed between the amygdala and emotion regulation Shahrokh et al., 2010). The OT-dopamine interactions are regions (Riem et al., 2012). Conversely, in response to an brain research 1580 (2014) 219–232 225

unknown infant's cry, intranasal OT decreased the women's experiences. The resulting alterations in OT receptor density, negative emotional arousal, as reflected by reduced amygdala affinity, or functions may underlie the decreased responsivity signals, while increasing activations in the empathy-related that is seen in these individuals upon exogenous adminis- regions (Riem et al., 2011). It is important to note that trations of OT (Bakermans-Kranenburg and van IJzendoorn, intranasal OT further decreased the women's handgrip force 2013). in response to hearing infant's cry, although the effect was This is of particular relevance in considering the use of OT present only in women without early experiences of harsh in psychiatric patients, since early adverse experiences are parenting (Bakermans-Kranenburg et al., 2012). rather common in this population. Not surprisingly, a review of data suggests that psychiatric patients have produced 5.3. Oxytocin in the treatment of psychopathology variable results in response to exogenous OT administrations, ranging from improvement (Guastella et al., 2010; Pedersen In addition to its role in regulating the expression of maternal et al., 2011) to worsening of symptoms (Bartz et al., 2011a; behavior as reviewed above, a large body of literature has Mah et al., 2013), along with some null findings (Epperson implicated OT in a much broader range of social behaviors et al., 1996; Pitman et al., 1993). In some cases, exogenous (Benarroch, 2013; Meyer-Lindenberg et al., 2011), including pair administrations of OT have yielded opposing patterns of bonding (Ross et al., 2009b; Schneiderman et al., 2012), inter- results for psychiatric and healthy groups (e.g., Bartz et al., personal trust (Kosfeld et al., 2005; Van IJzendoorn and 2011a; Pincus et al., 2010). This divergence is well reflected in Bakermans-Kranenburg, 2012), emotion recognition (Lischke the results of recent meta-analyses of available clinical trials, et al., 2012; Perry et al., 2013), and empathy (Hurlemann et al., which demonstrated that, when taken together, exogenous 2010; Rodrigues et al., 2009), to name a few. Due to its seemingly OT administrations did not improve symptoms of psy- widespread prosocial effects, along with its antidepressant and chiatric disorders with the exception of autism spectrum anxiolytic properties, OT has gained widespread popularity over disorder, although weak to moderate beneficial effects were the past decade in the study of normative and psychiatric found for healthy controls (Bakermans-Kranenburg and van populations. A sizable number of trials have investigated its IJzendoorn, 2013). While meta-analytic results are discoura- therapeutic potential in many psychiatric disorders, and many ging, more research is necessary given that the number of more trials are underway. To date, clinical trials in which OT clinical trials available for the meta-analyses were small for has demonstrated therapeutic benefitoverplaceboinclude many psychiatric conditions. those of autism spectrum disorder (Anagnostou et al., 2012; Andari et al., 2010; Guastella et al., 2010), schizophrenia 5.4. Can exogenous oxytocin benefit the treatment of (Averbeck et al., 2012; Feifel et al., 2010; Modabbernia et al., postpartum depression? 2013; Pedersen et al., 2011), social anxiety (Guastella et al., 2009; Labuschagne et al., 2010), and post-traumatic stress disorder While data from animal models suggest that OT may have (Yatzkar and Klein, 2009). potential in the treatment of PPD, the future of exogenous OT Despite the considerable excitement that these results in human psychiatric disorders remains unclear. To date, have generated, a more complicated and nuanced picture limited data exist on the antidepressant or anxiolytic effects emerges upon careful examination of the available data. of exogenous OT in women, and a small number of available Effects of OT reported in many social domains (e.g., social studies have demonstrated null (Rupp et al., 2012) or negative cognition, prosociality) are often inconsistent or, more pre- results (Domes et al., 2010; Mah et al., 2013). The role of cisely, are moderated by contextual and personal factors exogenous OT in human mothering has received more direct (Bartz et al., 2011b; Guastella and MacLeod, 2012; Macdonald support (Riem et al., 2011, 2012), although growing evidence and Macdonald, 2010). The context-dependent nature of the suggests that its effects are critically moderated by women's effects of OT is well demonstrated in a series of studies early caregiving experiences (Bakermans-Kranenburg et al., conducted on trust, in which exogenous administrations of 2012). While we are cautiously optimistic about OT's ther- OT increased participants' trust of individuals perceived as apeutic potential, we believe that there are many questions part of the in-group, but increased non-cooperation toward that remain to be answered. Here, we highlight some of these out-group members who were perceived as potential threats remaining questions for future research. (De Dreu et al., 2010, 2012). The person-dependent nature of First, it would be of great importance to identify individual the effects of OT is well captured in a growing number of differences among PPD patients that may moderate the effects studies that demonstrate null effects of exogenous OT in of exogenous OT. PPD patients are likely a heterogeneous individuals with adverse early caregiving experiences, group of individuals demonstrating a large variability in OT whether assessed by reported severity of childhood abuse, receptor function, endogenous OT production, and early car- neglect, or loss (Meinlschmidt and Heim, 2007), memories of egiving experiences. As these factors are understood to alter parental love-withdrawal (Riem et al., 2013; van Ijzendoorn one's responsivity to exogenous OT, careful investigation of et al., 2011), or recollections of harsh parenting experiences these intraindividual characteristics and identification of rele- (Bakermans-Kranenburg et al., 2012). Considering that the vant neurobiological and behavioral markers are of paramount development of one's OT system may be critically modified interest. Given the pattern of results reviewed, it is possible by the quality of early caregiving one receives (Champagne, that exogenous OT may yield beneficial effects only in a 2008; Meaney, 2001), it is possible that the OT system may subgroup of PPD patients. Such results may not be apparent have been altered at a more fundamental level, possibly at and may even be obscured in between-group designs, where the level of receptors, in individuals with early adverse effects are averaged across individuals and within-group 226 brain research 1580 (2014) 219–232

individual differences are overlooked (Guastella and MacLeod, variations are necessary to fully evaluate the therapeutic 2012). We concur with others (e.g., Guastella and MacLeod, 2012) potential of OT in the treatment of PPD. that the imperative next step in OTtranslationalresearchisto develop cognitive, behavioral, and neurobiological markers that can index the degree to which patients may be responsive to Funding sources exogenous administrations of OT. Second, gender is an important consideration given the This work was supported by the Eunice Kennedy Shriver differences in the endogenous OT levels between men and National Institute of Child Health and Human Development women. Furthermore, gonadal hormones, estrogen in parti- [R01HD065819]; and the National Institute on Drug Abuse cular, are understood to be critically involved in the regula- [R01DA026437]. The content is solely the responsibility of the tion of OT, whether endogenously produced or exogenously authors and does not necessarily represent the official views administered. It is in this regard that a relative dearth of of these institutes or the National Institutes of Health. clinical trials in women is particularly problematic. It is not fi yet clear to what degree the extant ndings obtained in men references can be generalized to women, and particularly to postpartum women who undergo significant hormonal changes. It is also unclear how, and to what extent, menstrual cycle and Abou-Saleh, M.T., et al., 1998. Hormonal aspects of postpartum fl accompanying uctuations in gonadal hormones moderate depression. Psychoneuroendocrinology 23, 465–475. the effects of exogenously administered OT and contribute to Ahokas, A., et al., 2001. Estrogen deficiency in severe postpartum divergent results (e.g., see Domes et al., 2010 vs. Rupp et al., depression: successful treatment with sublingual physiologic 2012). It is important that future studies recruit female 17beta-estradiol: a preliminary study. J. Clin. Psychiatry 62, participants, and specifically postpartum women, to examine 332–336. Alpern, L., Lyons-Ruth, K., 1993. Preschool children at social risk: the effects of OT in this unique population. chronicity and timing of maternal depressive symptoms and Third, for therapeutic use of exogenous OT, it would be child behavior problems at school and at home. Dev. important to systematically address questions of dosage, Psychopathol 5, 371–387. timing, and side effects associated with long-term adminis- Altemus, M., et al., 1995. Suppression of hypothalmic–pituitary– tration. All clinical trials to date have used low doses (18 to adrenal axis responses to stress in lactating women. J. Clin. 40 IU) for a short span of time, with minimal reported side Endocrinol. Metab 80, 2954–2959. effects (MacDonald et al., 2011). Future research should American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Mental Disorders, fourth ed. American examine the safety of high-dose long-term use of OT. It Psychiatric Association, Washington, DC (text rev.). would be critical to understand how chronic exogenous American Psychiatric Association, 2013. Diagnostic and administrations of OT may affect endogenous OT production Statistical Manual of Mental Disorders, fifth ed. American as well as complex neuroendocrine functions in the post- Psychiatric Publishing, Arlington, VA. partum period. Amico, J.A., Johnston, J.M., Vagnucci, A.H., 1994. Suckling-induced Fourth, it remains to be determined whether exogenous attenuation of plasma cortisol concentrations in postpartum OT should be used as a stand-alone treatment or should lactating women. Endocrine Res. 20, 79–87. Amico, J.A., et al., 2008. Oxytocin knockout mice: a model for better be integrated with other therapies. Most available studying stress-related and ingestive behaviours. Prog. Brain clinical trials have excluded patients undergoing medication Res. 170, 53–64. treatment, and there is currently limited data from which to Anagnostou, E., et al., 2012. Intranasal oxytocin versus placebo in draw conclusions about the safety and efficacy of OT as an the treatment of adults with autism spectrum disorders: a augmentation agent (MacDonald et al., 2011). Furthermore, randomized controlled trial. Mol. Autism 3, 16. anxiolytic and prosocial effects of OT have led some to Andari, E., et al., 2010. Promoting social behavior with oxytocin in high-functioning autism spectrum disorders. Proc. Nat. Acad. believe that exogenous OT administrations may help aid Sci. U.S.A. 107, 4389–4394. the process of psychotherapy. This would be a fruitful area Appleby, L., et al., 1997. A controlled study of fluoxetine and for future investigation. cognitive-behavioural counselling in the treatment of postnatal depression. BMJ 314, 932–936. Apter-Levy, Y., et al., 2013. Impact of maternal depression across the first 6 years of life on the child’s mental health, social 6. Conclusions engagement, and empathy: the moderating role of oxytocin. Am. J. Psychiatry 170, 1161–1168. Arletti, R., Bertolini, A., 1987. Oxytocin acts as an antidepressant While OT initially appeared to offer much promise, the in two animal models of depression. Life Sci. 41, 1725–1730. pattern of results that has thus far emerged is more nuanced Atzil, S., Hendler, T., Feldman, R., 2011. Specifying the and inconsistent than it appeared to be at first. The literature neurobiological basis of human attachment: brain, hormones, reviewed in preceding sections suggest that studies are and behavior in synchronous and intrusive mothers. beginning to shed light upon the complex context- and Neuropsychopharmacology 36, 2603–2615. person-dependent nature of OT effects. We propose that Averbeck, B.B., et al., 2012. Emotion recognition and oxytocin in patients with schizophrenia. Psychol. Med. 42, 259–266. future studies attend to individual variations that may be Azak, S., Raeder, S., 2013. Trajectories of parenting behavior and present among mothers with PPD, rather than looking for maternal depression. Infant Behav. Dev 36, 391–402. the uniform effect across all mothers. We underscore that Bakermans-Kranenburg, M.J., et al., 2012. Oxytocin decreases focused studies that tease apart OT-related individual handgrip force in reaction to infant crying in females without brain research 1580 (2014) 219–232 227

harsh parenting experiences. Soc. Cogn. Affect Neurosci 7, Champagne, F.A., 2008. Epigenetic mechanisms and the 951–957. transgenerational effects of maternal care. Front. Bakermans-Kranenburg, M.J., van IJzendoorn, M.H., 2013. Sniffing Neuroendocrinol. 29, 386–397. around oxytocin: review and meta-analyses of trials in Chiodera, P., et al., 1991. Relationship between plasma profiles of healthy and clinical groups with implications for oxytocin and adrenocorticotropic hormone during suckling or pharmacotherapy. Transl. Psychiatry 3, e258. breast stimulation in women. Horm. Res. 35, 119–123. Bartz, J., et al., 2011a. Oxytocin can hinder trust and cooperation Cicchetti, D., Rogosch, F.A., Toth, S.L., 2000. The efficacy of in borderline personality disorder. Soc. Cogn. Affect. Neurosci toddler-parent psychotherapy for fostering cognitive 6, 556–563. development in offspring of depressed mothers. J. Abnorm. Bartz, J.A., Hollander, E., 2006. The neuroscience of affiliation: Child Psychol. 28, 135–148. forging links between basic and clinical research on Cipriani, A., et al., 2007. Metareview on short-term effectiveness neuropeptides and social behavior. Horm. Behav 50, 518–528. and safety of antidepressants for depression: an evidence- Bartz, J.A., et al., 2011b. Social effects of oxytocin in humans: based approach to inform clinical practice. Can. J. Psychiatry context and person matter. Trends Cogn. Sci 15, 301–309. 52, 553–562. Beck, C.T., 2001. Predictors of postpartum depression: an update. Clark, R., Tluczek, A., Wenzel, A., 2003. Psychotherapy for Nurs. Res. 50, 275–285. postpartum depression: a preliminary report. Am. J. Beck, C.T., 2006. Postpartum depression: it isn’t just the blues. Orthopsychiatry 73, 441–454. Am. J. Nurs. 106, 40–50. Cooper, P.J., et al., 2003. Controlled trial of the short- and long- Benarroch, E.E., 2013. Oxytocin and vasopressin: social term effect of psychological treatment of post-partum neuropeptides with complex neuromodulatory functions. depression. I. Impact on maternal mood. Br. J. Psychiatry 182, Neurology 80, 1521–1528. 412–419. Bennett, H.A., et al., 2004. Depression during pregnancy: overview Corwin, E.J., et al., 2010. The heritability of postpartum of clinical factors. Clin. Drug Investig 24, 157–179. depression. Biol. Res. Nurs. 12, 73–83. Bledsoe, S.E., Grote, N.K., 2006. Treating depression during Cuijpers, P., Brannmark, J.G., van Straten, A., 2008. Psychological pregnancy and the postpartum: a preliminary meta-analysis. treatment of postpartum depression: a meta-analysis. J. Clin. Res. Social Work Pract. 16, 109–120. Psychol. 64, 103–118. Bloch, M., et al., 2000. Effects of gonadal steroids in women with a Cyranowski, J.M., et al., 2008. Evidence of dysregulated peripheral history of postpartum depression. Am. J. Psychiatry 157, oxytocin release among depressed women. Psychosom. Med. 924–930. 70, 967–975. Bloch, M., et al., 2005. Risk factors associated with the Dawes, M., Chowienczyk, P.J., 2001. Drugs in pregnancy. development of postpartum mood disorders. J. Affect. Pharmacokinetics in pregnancy. Best Pract. Res. Clin. Obstet. Disorders 88, 9–18. Gynaecol 15, 819–826. Bloch, M., et al., 2012. The effect of sertraline add-on to brief Dawson, G., et al., 2003. Preschool outcomes of children of dynamic psychotherapy for the treatment of postpartum depressed mothers: role of maternal behavior, contextual risk, depression: a randomized, double-blind, placebo-controlled and children’s brain activity. Child Dev. 74, 1158–1175. study. J. Clin. Psychiatry 73, 235–241. De Crescenzo, F., et al., 2013. Selective serotonin reuptake Blom, E.A., et al., 2010. Perinatal complications increase the risk inhibitors (SSRIs) for post-partum depression (PPD): a of postpartum depression. The generation R study. BJOG 117, systematic review of randomized clinical trials. J. Affect. 1390–1398. Disorders, http://dxdoi.org/10.1016/j.jad.2013.09.019. Buttner, M.M., O’Hara, M.W., Watson, D., 2012. The structure of De Dreu, C.K., et al., 2010. The neuropeptide oxytocin regulates women’s mood in the early postpartum. Assessment 19, parochial altruism in intergroup conflict among humans. 247–256. Science 328, 1408–1411. Chabrol, H., et al., 2002. Prevention and treatment of post-partum De Dreu, C.K., et al., 2012. Oxytocin modulates selection of allies depression: a controlled randomized study on women at risk. in intergroup conflict. Proc. Biol. Sci 279, 1150–1154. Psychol. Med. 32, 1039–1047. Demitrack, M.A., Gold, P.W., 1988. Oxytocin: neurobiologic Chabrol, H., et al., 2004. Acceptability of psychotherapy and considerations and their implications for affective illness. antidepressants for postnatal depression among newly Prog. Neuro-Psychopharmacol. Biol. Psychiatry 12 (Suppl.), delivered mothers. J. Reprod. Infant Psychol. 22, 5–12. S23–51. Champagne, F., et al., 2001. Naturally occurring variations in Dennis, C.L., Janssen, P.A., Singer, J., 2004. Identifying women at- maternal behavior in the rat are associated with differences in risk for postpartum depression in the immediate postpartum estrogen-inducible central oxytocin receptors. Proc. Natl. period. Acta Psychiatr. Scand. 110, 338–346. Acad. Sci. USA 98, 12736–12741. Dennis, C.L., Hodnett, E., 2007. Psychosocial and psychological Champagne, F.A., et al., 2003a. Variations in maternal care in the interventions for treating postpartum depression. Cochrane rat as a mediating influence for the effects of environment on Database Syst. Rev (CD006116). development. Physiol. Behav. 79, 359–371. Dennis, C.L., McQueen, K., 2007. Does maternal postpartum Champagne, F.A., et al., 2003b. Natural variations in maternal depressive symptomatology influence infant feeding care are associated with estrogen receptor alpha expression outcomes?. Acta Paediatr 96, 590–594. and estrogen sensitivity in the medial preoptic area. Dennis, C.L., Ross, L.E., Herxheimer, A., 2008. Oestrogens and Endocrinology 144, 4720–4724. progestins for preventing and treating postpartum depression. Champagne, F.A., et al., 2004. Variations in nucleus accumbens Cochrane Database Syst. Rev (CD001690). dopamine associated with individual differences in maternal Dennis, C.L., McQueen, K., 2009. The relationship between infant- behavior in the rat. J. Neurosci. 24, 4113–4123. feeding outcomes and postpartum depression: a qualitative Champagne, F.A., et al., 2006. Maternal care associated with systematic review. Pediatrics 123, e736–e751. methylation of the estrogen receptor-alpha1b promoter and Ditzen, B., et al., 2009. Intranasal oxytocin increases positive estrogen receptor-alpha expression in the medial preoptic communication and reduces cortisol levels during couple area of female offspring. Endocrinology 147, 2909–2915. conflict. Biol. Psychiatry 65, 728–731. 228 brain research 1580 (2014) 219–232

Domes, G., et al., 2010. Effects of intranasal oxytocin on Forty, L., et al., 2006. Familiality of postpartum depression in emotional face processing in women. unipolar disorder: results of a family study. Am. J. Psychiatry Psychoneuroendocrinology 35, 83–93. 163, 1549–1553. Doornbos, B., et al., 2009. The development of peripartum Francis, D., et al., 1999. Nongenomic transmission across depressive symptoms is associated with gene polymorphisms generations of maternal behavior and stress responses in the of MAOA, 5-HTT and COMT. Prog. Neuro-Psychopharmacol. rat. Science 286, 1155–1158. Biol. Psychiatry 33, 1250–1254. Francis, D.D., Champagne, F.C., Meaney, M.J., 2000. Variations in Ebner, K., et al., 2000. A single social defeat experience selectively maternal behaviour are associated with differences in stimulates the release of oxytocin, but not vasopressin, within oxytocin receptor levels in the rat. J. Neuroendocrinol. 12, the septal brain area of male rats. Brain Res. 872, 87–92. 1145–1148. Ebner, K., et al., 2005. Release of oxytocin in the rat central Frasch, A., et al., 1995. Reduction of plasma oxytocin levels in amygdala modulates stress-coping behavior and the release of patients suffering from major depression. Adv. Exp. Med. Biol excitatory amino acids. Neuropsychopharmacology 30, 223–230. 395, 257–258. Ellfolk, M., Malm, H., 2010. Risks associated with in utero and Gavin, N.I., et al., 2005. Perinatal depression: a systematic review lactation exposure to selective serotonin reuptake inhibitors of prevalence and incidence. Obstetr. Gynecol 106, 1071–1083. (SSRIs). Reprod. Toxicol 30, 249–260. Gaynes, B.N., et al., 2005. Perinatal Depression: Prevalence, Elliott, S., 2000. Report on the Satra Bruk workshop on Screening Accuracy, and Screening Outcomes. Evidence classification of postnatal mental disorders. Arch. Womens Report/Technology Assessment No. 119 (Prepared by the RTI- Ment. Health 3, 27–33. University of North Carolina Evidence-based Practice Center, Emiliano, A., et al., 2007. The interface of oxytocin-labeled cells under Contract No. 290-02-0016.) AHRQ Publication and serotonin transporter-containing fibers in the primate No. 05-E006-2. hypothalamus: a substrate for SSRIs therapeutic effects?. Gimpl, G., Fahrenholz, F., 2001. The oxytocin receptor system: Neuropsychopharmacology 32, 977–988. structure, function, and regulation. Physiol. Rev. 81, 629–683. Engelmann, M., Landgraf, R., Wotjak, C.T., 2004. The Gordon, I., et al., 2008. Oxytocin and cortisol in romantically hypothalamic-neurohypophysial system regulates the unattached young adults: associations with bonding and hypothalamic–pituitary–adrenal axis under stress: an old psychological distress. Psychophysiology 45, 349–352. concept revisited. Front. Neuroendocrinol 25, 132–149. Gordon, I., et al., 2010a. Oxytocin, cortisol, and triadic family Epperson, C.N., McDougle, C.J., Price, L.H., 1996. Intranasal interactions. Physiol. Behav. 101, 679–684. oxytocin in obsessive–compulsive disorder. Biol. Psychiatry 40, Gordon, I., et al., 2010b. Oxytocin and the development of 547–549. parenting in humans. Biol. Psychiatry 68, 377–382. Epperson, C.N., et al., 2003. Maternal fluoxetine treatment in the Gregoire, A.J., et al., 1996. Transdermal oestrogen for treatment of postpartum period: effects on serotonin and plasma severe postnatal depression. Lancet 347, 930–933. drug levels in breastfeeding mother-infant pairs. Pediatrics Groer, M.W., Morgan, K., 2007. Immune, health and endocrine 112, e425. characteristics of depressed postpartum mothers. Feifel, D., et al., 2010. Adjunctive intranasal oxytocin reduces Psychoneuroendocrinology 32, 133–139. symptoms in schizophrenia patients. Biol. Psychiatry 68, Grote, N.K., et al., 2009. A randomized controlled trial of culturally 678–680. relevant, brief interpersonal psychotherapy for perinatal Feldman, R., 2007. Maternal versus child risk and the depression. Psychiatr. Serv. 60, 313–321. development of parent–child and family relationships in five Guastella, A.J., et al., 2009. A randomized controlled trial of high-risk populations. Dev. Psychopathol 19, 293–312. intranasal oxytocin as an adjunct to exposure therapy for social Feldman, R., et al., 2007. Evidence for a neuroendocrinological anxiety disorder. Psychoneuroendocrinology 34, 917–923. foundation of human affiliation: plasma oxytocin levels Guastella, A.J., et al., 2010. Intranasal oxytocin improves emotion across pregnancy and the postpartum period predict mother– recognition for youth with autism spectrum disorders. Biol. infant bonding. Psychol. Sci 18, 965–970. Psychiatry 67, 692–694. Feldman, R., et al., 2010a. Natural variations in maternal and Guastella, A.J., MacLeod, C., 2012. A critical review of the paternal care are associated with systematic changes in influence of oxytocin nasal spray on social cognition in oxytocin following parent–infant contact. humans: evidence and future directions. Horm. Behav. 61, Psychoneuroendocrinology 35, 1133–1141. 410–418. Feldman, R., Gordon, I., Zagoory-Sharon, O., 2010b. The cross- Harris, B., et al., 1996. Cardiff puerperal mood and hormone generation transmission of oxytocin in humans. Horm. Behav. study. III. Postnatal depression at 5 to 6 weeks postpartum, 58, 669–676. and its hormonal correlates across the peripartum period. Feldman, R., Gordon, I., Zagoory-Sharon, O., 2011. Maternal and Br. J. Psychiatry 168, 739–744. paternal plasma, salivary, and urinary oxytocin and parent- Hay, D.F., et al., 2001. Intellectual problems shown by 11-year-old infant synchrony: considering stress and affiliation children whose mothers had postnatal depression. J. Child components of human bonding. Dev. Sci 14, 752–761. Psychol. Psychiatry Allied Disciplines 42, 871–889. Field, T., 2010. Postpartum depression effects on early Heim, C., et al., 2009. Lower CSF oxytocin concentrations in women interactions, parenting, and safety practices: a review. Infant with a history of childhood abuse. Mol. Psychiatry 14, 954–958. Behav. Dev 33, 1–6. Heinrichs, M., et al., 2001. Effects of suckling on hypothalamic– Fitelson, E., et al., 2010. Treatment of postpartum depression: pituitary–adrenal axis responses to psychosocial stress in clinical, psychological and pharmacological options. Int. J. postpartum lactating women. J. Clin. Endocrinol. Metab. 86, Womens Health 3, 1–14. 4798–4804. Flynn, H.A., et al., 2004. Rates of maternal depression in pediatric Heinrichs, M., et al., 2003. Social support and oxytocin interact to emergency department and relationship to child service suppress cortisol and subjective responses to psychosocial utilization. Gen. Hospital Psychiatry 26, 316–322. stress. Biol. Psychiatry 54, 1389–1398. Forman, D.R., et al., 2007. Effective treatment for postpartum Hendrick, V., et al., 2000. Postpartum and nonpostpartum depression is not sufficient to improve the developing depression: differences in presentation and response to mother–child relationship. Dev. Psychopathol 19, 585–602. pharmacologic treatment. Depress Anxiety 11, 66–72. brain research 1580 (2014) 219–232 229

Hinshaw, K., et al., 2008. A randomised controlled trial of early Lightman, S.L., et al., 2001. Peripartum plasticity within the versus delayed oxytocin augmentation to treat primary hypothalamo–pituitary–adrenal axis. Prog. Brain Res 133, dysfunctional labour in nulliparous women. BJOG 115, 111–129. 1289–1295. Lischke, A., et al., 2012. Intranasal oxytocin enhances emotion Hurlemann, R., et al., 2010. Oxytocin enhances amygdala- recognition from dynamic facial expressions and leaves eye- dependent, socially reinforced learning and emotional gaze unaffected. Psychoneuroendocrinology 37, 475–481. empathy in humans. J. Neurosci 30, 4999–5007. Logsdon, M.C., Wisner, K., Hanusa, B.H., 2009. Does maternal role Insel, T.R., Harbaugh, C.R., 1989. Lesions of the hypothalamic functioning improve with antidepressant treatment in women paraventricular nucleus disrupt the initiation of maternal with postpartum depression?. J. Womens Health (Larchmt) 18, behavior. Physiol. Behav 45, 1033–1041. 85–90. Insel, T.R., 2010. The challenge of translation in social Logsdon, M.C., et al., 2011. Depression treatment and maternal neuroscience: a review of oxytocin, vasopressin, and affiliative functioning. Depress. Anxiety 28, 1020–1026. behavior. Neuron 65, 768–779. Lommatzsch, M., et al., 2006. Maternal serum concentrations of Jameson, P.B., et al., 1997. Mother-toddler interaction patterns BDNF and depression in the perinatal period. associated with maternal depression. Dev. Psychopathol 9, Psychoneuroendocrinology 31, 388–394. 537–550. Lovejoy, M.C., et al., 2000. Maternal depression and parenting Javed, A., et al., 1999. D-Fenfluramine induces serotonin-mediated behavior: a meta-analytic review. Clin. Psychol. Rev. 20, Fos expression in corticotropin-releasing factor and oxytocin 561–592. neurons of the hypothalamus, and serotonin-independent Fos Lumley, J., Austin, M.P., Mitchell, C., 2004. Intervening to reduce expression in enkephalin and neurotensin neurons of the depression after birth: a systematic review of the randomized amygdala. Neuroscience 90, 851–858. trials. Int. J. Technol. Assess. Health Care 20, 128–144. Jolley, S.N., et al., 2007. Dysregulation of the hypothalamic– MacDonald, E., et al., 2011. A review of safety, side-effects and pituitary–adrenal axis in postpartum depression. Biol. Res. subjective reactions to intranasal oxytocin in human Nurs 8, 210–222. research. Psychoneuroendocrinology 36, 1114–1126. Kammerer, M., Taylor, A., Glover, V., 2006. The HPA axis and Macdonald, K., Macdonald, T.M., 2010. The peptide that binds: a perinatal depression: a hypothesis. Arch. Womens Ment. systematic review of oxytocin and its prosocial effects in Health 9, 187–196. humans. Harvard Rev. Psychiatry 18, 1–21. Kendell, R.E., Chalmers, J.C., Platz, C., 1987. Epidemiology of Mah, B.L., et al., 2013. Oxytocin in postnatally depressed mothers: puerperal psychoses. Br. J. Psychiatry 150, 662–673. its influence on mood and expressed emotion. Prog. Neuro- Kendrick, K.M., Keverne, E.B., Baldwin, B.A., 1987. Psychopharmacol. Biol. Psychiatry 40, 267–272. Intracerebroventricular oxytocin stimulates maternal Malphurs, J.E., et al., 1996. Touch by intrusive and withdrawn behaviour in the sheep. Neuroendocrinology 46, 56–61. mothers with depressive symptoms. Early Dev. Parenting 5, Keverne, E.B., Kendrick, K.M., 1992. Oxytocin facilitation of 111–115. maternal behavior in sheep. Ann. N. Y. Acad. Sci 652, 83–101. McCarthy, M.M., 1990. Oxytocin inhibits infanticide in Kim, D.R., O’Reardon, J.P., Epperson, C.N., 2010. Guidelines for the female house mice (Mus domesticus). Horm. Behav 24, management of depression during pregnancy. Curr. Psychiatry 365–375. Rep. 12, 279–281. McLearn, K.T., et al., 2006. Maternal depressive symptoms at 2 to Kim, P., et al., 2011. Breastfeeding, brain activation to own infant 4 months post partum and early parenting practices. Arch. cry, and maternal sensitivity. J. Child Psychol. Psychiatry Pediatr. Adolescent Med 160, 279–284. Allied Disciplines 52, 907–915. Meaney, M.J., 2001. Maternal care, gene expression, and the Kim, S., et al., 2014. Maternal oxytocin response predicts mother-to- transmission of individual differences in stress reactivity infant gaze. Brain Res. 1580, 133–142, http://dx.doi.org/10.1016/j. across generations. Ann. Rev. Neurosci 24, 1161–1192. brainres.2013.10.050. Meinlschmidt, G., Heim, C., 2007. Sensitivity to intranasal Kirsch, I., et al., 2008. Initial severity and antidepressant benefits: oxytocin in adult men with early parental separation. Biol. a meta-analysis of data submitted to the Food and Drug Psychiatry 61, 1109–1111. Administration. PLoS Med 5, e45. Melis, M.R., et al., 2007. Oxytocin injected into the ventral Kirsch, P., et al., 2005. Oxytocin modulates neural circuitry for tegmental area induces penile erection and increases social cognition and fear in humans. J. Neurosci 25, extracellular dopamine in the nucleus accumbens and 11489–11493. paraventricular nucleus of the hypothalamus of male rats. Klier, C.M., et al., 2007. The role of estrogen and progesterone in Eur. J. Neurosci 26, 1026–1035. depression after birth. J. Psychiatr. Res 41, 273–279. Meyer-Lindenberg, A., et al., 2011. Oxytocin and vasopressin in Kosfeld, M., et al., 2005. Oxytocin increases trust in humans. the human brain: social neuropeptides for translational Nature 435, 673–676. medicine. Nat. Rev. Neurosci 12, 524–538. Labuschagne, I., et al., 2010. Oxytocin attenuates amygdala Milgrom, J., et al., 2005. A randomized controlled trial of reactivity to fear in generalized social anxiety disorder. psychological interventions for postnatal depression. Br. J. Neuropsychopharmacology 35, 2403–2413. Clin. Psychol 44, 529–542. Lancel, M., Kromer, S., Neumann, I.D., 2003. Intracerebral Mills-Koonce, W.R., et al., 2008. Changes in maternal sensitivity oxytocin modulates sleep-wake behaviour in male rats. Regul. across the first three years: are mothers from different Pept 114, 145–152. attachment dyads differentially influenced by depressive Laurent, H.K., Ablow, J.C., 2013. A face a mother could love: symptomatology?. Attach. Hum. Dev 10, 299–317. depression-related maternal neural responses to infant Misri, S., et al., 2004. Restoration of functionality in postpartum emotion faces. Soc. Neurosci 8, 228–239. depressed mothers: an open-label study with escitalopram. Li, L., et al., 1999. Regulation of maternal behavior and offspring J. Clin. Psychopharmacol 32, 729–732. growth by paternally expressed Peg3. Science 284, 330–333. Modabbernia, A., et al., 2013. Intranasal oxytocin as an adjunct to Liberzon, I., et al., 1997. Motivational properties of oxytocin risperidone in patients with schizophrenia: an 8-week, in the conditioned place preference paradigm. randomized, double-blind, placebo-controlled study. CNS Neuropsychopharmacology 17, 353–359. Drugs 27, 57–65. 230 brain research 1580 (2014) 219–232

Moses-Kolko, E.L., et al., 2009. Transdermal estradiol for Paulson, J.F., Dauber, S., Leiferman, J.A., 2006. Individual and postpartum depression: a promising treatment option. Clin. combined effects of postpartum depression in mothers and Obstetr. Gynecol 52, 516–529. fathers on parenting behavior. Pediatrics 118, 659–668. Mulcahy, R., et al., 2010. A randomised control trial for the Payne, J.L., et al., 2008. Familial aggregation of postpartum mood effectiveness of group Interpersonal Psychotherapy for symptoms in bipolar disorder pedigrees. Bipolar Disord 10, postnatal depression. Arch. Womens Ment. Health 13, 125–139. 38–44. Munk-Olsen, T., et al., 2006. New parents and mental disorders: a Pearlstein, T.B., et al., 2006. Patient choice of treatment for population-based register study. JAMA 296, 2582–2589. postpartum depression: a pilot study. Arch. Womens Ment. Murphy-Eberenz, K., et al., 2006. Is perinatal depression familial?. Health 9, 303–308. J. Affect. Disorders 90, 49–55. Pedersen, C.A., Prange Jr., A.J., 1979. Induction of maternal Murray, L., Cooper, P.J., 1997. Postpartum depression and child behavior in virgin rats after intracerebroventricular development. Psychol. Med. 27, 253–260. administration of oxytocin. Proc. Nat. Acad. Sci. U.S.A 76, Murray, L., et al., 2003. Controlled trial of the short- and long-term 6661–6665. effect of psychological treatment of post-partum depression: Pedersen, C.A., et al., 1982. Oxytocin induces maternal behavior 2. Impact on the mother–child relationship and child in virgin female rats. Science 216, 648–650. outcome. Br. J. Psychiatry 182, 420–427. Pedersen, C.A., et al., 1994. Oxytocin activates the postpartum Murray, L., et al., 2011. Maternal postnatal depression and the onset of rat maternal behavior in the ventral tegmental and development of depression in offspring up to 16 years of age. medial preoptic areas. Behav. Neurosci 108, 1163–1171. J. Am. Acad. Child Adolescent Psychiatry 50, 460–470. Pedersen, C.A., et al., 2011. Intranasal oxytocin reduces psychotic Neumann, I.D., Torner, L., Wigger, A., 2000. Brain oxytocin: symptoms and improves Theory of Mind and social differential inhibition of neuroendocrine stress responses and perception in schizophrenia. Schizophrenia Res 132, 50–53. anxiety-related behaviour in virgin, pregnant and lactating Perry, A., et al., 2013. Face or body? Oxytocin improves perception rats. Neuroscience 95, 567–575. of emotions from facial expressions in incongruent emotional Neumann, I.D., 2003. Brain mechanisms underlying emotional body context. Psychoneuroendocrinology 38, 2820–2825. alterations in the peripartum period in rats. Depress. Anxiety Pincus, D., et al., 2010. Inverse effects of oxytocin on attributing mental activity to others in depressed and healthy subjects: a 17, 111–121. double-blind placebo controlled FMRI study. Front. Psychiatry Neumann, I.D., Landgraf, R., 2012. Balance of brain oxytocin and 1, 134. vasopressin: implications for anxiety, depression, and social Pitman, R.K., Orr, S.P., Lasko, N.B., 1993. Effects of intranasal behaviors. Trends Neurosci. 35, 649–659. vasopressin and oxytocin on physiologic responding during NICHD Early Child Care Research Network, 1999. Chronicity of personal combat imagery in Vietnam veterans with maternal depressive symptoms, maternal sensitivity, and posttraumatic stress disorder. Psychiatry Res 48, 107–117. child functioning at 36 months.. Dev. Psychol. 35, 1297-1310. Pitts, A.F., et al., 1995. Cerebrospinal fluid corticotropin-releasing Nishioka, T., et al., 1998. Stress increases oxytocin release within hormone, vasopressin, and oxytocin concentrations in treated the hypothalamic paraventricular nucleus. Brain Res 781, patients with major depression and controls. Biol. Psychiatry 57–61. 38, 330–335. Nomura, M., et al., 2003. Enhanced up-regulation of corticotropin- Poobalan, A.S., et al., 2007. Effects of treating postnatal releasing hormone gene expression in response to restraint depression on mother-infant interaction and child stress in the hypothalamic paraventricular nucleus of development: systematic review. Br. J. Psychiatry 191, oxytocin gene-deficient male mice. J. Neuroendocrinol 15, 378–386. 1054–1061. Ragnauth, A.K., et al., 2005. Female oxytocin gene-knockout mice, Nowakowska, E., et al., 2002. Role of neuropeptides in in a semi-natural environment, display exaggerated antidepressant and memory improving effects of venlafaxine. aggressive behavior. Genes Brain Behav 4, 229–239. Pol. J. Pharmacol 54, 605–613. Riem, M.M., et al., 2011. Oxytocin modulates amygdala, insula, Numan, M., Smith, H.G., 1984. Maternal behavior in rats: evidence and inferior frontal gyrus responses to infant crying: a for the involvement of preoptic projections to the ventral randomized controlled trial. Biol. Psychiatry 70, 291–297. tegmental area. Behav. Neurosci 98, 712–727. Riem, M.M., et al., 2012. No laughing matter: intranasal oxytocin O’Hara, M.W., Swain, A.M., 1996. Rates and risk of postpartum administration changes functional brain connectivity during depression—a meta-analysis. Int. Rev. Psychiatry 8, 37–54. exposure to infant laughter. Neuropsychopharmacology 37, O’Hara, M.W., et al., 2000. Efficacy of interpersonal psychotherapy 1257–1266. for postpartum depression. Arch. Gen. Psychiatry 57, Riem, M.M., et al., 2013. Does intranasal oxytocin promote 1039–1045. prosocial behavior to an excluded fellow player? A O’Hara, M.W., McCabe, J.E., 2013. Postpartum depression: current randomized-controlled trial with Cyberball. status and future directions. Annu. Rev. Clin. Psychol. 9, Psychoneuroendocrinology 38, 1418–1425. 379–407. Ring, R.H., et al., 2010. Receptor and behavioral of Okano, T., Nomura, J., 1992. Endocrine study of the maternity WAY-267464, a non-peptide oxytocin receptor . blues. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 16, Neuropharmacology 58, 69–77. 921–932. Robertson, E., et al., 2004. Antenatal risk factors for postpartum Olazabal, D.E., Young, L.J., 2006. Species and individual depression: a synthesis of recent literature. Gen. Hospital differences in juvenile female alloparental care are associated Psychiatry 26, 289–295. with oxytocin receptor density in the striatum and the lateral Rodrigues, S.M., et al., 2009. Oxytocin receptor genetic variation septum. Horm. Behav. 49, 681–687. relates to empathy and stress reactivity in humans. Proc. Nat. Ozsoy, S., Esel, E., Kula, M., 2009. Serum oxytocin levels in Acad. Sci. U.S.A. 106, 21437–21441. patients with depression and the effects of gender and Ross, H.E., et al., 2009a. Characterization of the oxytocin system antidepressant treatment. Psychiatry Res 169, 249–252. regulating affiliative behavior in female prairie voles. Pang, W.W., Hartmann, P.E., 2007. Initiation of human lactation: Neuroscience 162, 892–903. secretory differentiation and secretory activation. J. Mammary Ross, H.E., et al., 2009b. Variation in oxytocin receptor density in Gland Biol. Neoplasia 12, 211–221. the nucleus accumbens has differential effects on affiliative brain research 1580 (2014) 219–232 231

behaviors in monogamous and polygamous voles. J. Neurosci Sullivan, P.F., Neale, M.C., Kendler, K.S., 2000. Genetic 29, 1312–1318. epidemiology of major depression: review and meta-analysis. Ross, H.E., Young, L.J., 2009. Oxytocin and the neural mechanisms Am. J. Psychiatry 157, 1552–1562. regulating social cognition and affiliative behavior. Front. Takayanagi, Y., et al., 2005. Pervasive social deficits, but normal Neuroendocrinol. 30, 534–547. parturition, in oxytocin receptor-deficient mice. Proc. Nat. Ross, L.E., et al., 2003. Measurement issues in postpartum Acad. Sci. U.S.A. 102, 16096–16101. depression Part 1: Anxiety as a feature of postpartum Taveras, E.M., et al., 2003. Clinician support and psychosocial risk depression. Arch. Womens Ment. Health 6, 51–57. factors associated with breastfeeding discontinuation. Rupp, H.A., et al., 2012. Amygdala response to negative images in Pediatrics 112, 108–115. postpartum vs nulliparous women and intranasal oxytocin. Taylor, A., et al., 2009. Diurnal pattern of cortisol output in Soc. Cogn. Affect Neurosci http://dx.doi.org/10.1093/scan/ postnatal depression. Psychoneuroendocrinology 34, nss100. 1184–1188. Sanjuan, J., et al., 2008. Mood changes after delivery: role of the Treloar, S.A., et al., 1999. Genetic influences on post-natal serotonin transporter gene. Br. J. Psychiatry 193, 383–388. depressive symptoms: findings from an Australian twin Scantamburlo, G., et al., 2007. Plasma oxytocin levels and anxiety sample. Psychol. Med 29, 645–654. in patients with major depression. Psychoneuroendocrinology Tronick, E., Reck, C., 2009. Infants of depressed mothers. Harvard 32, 407–410. Rev. Psychiatry 17, 147–156. Schneiderman, I., et al., 2012. Oxytocin during the initial stages of Troutman, B.R., Cutrona, C.E., 1990. Nonpsychotic postpartum romantic attachment: relations to couples’ interactive depression among adolescent mothers. J. Abnorm. Psychol. reciprocity. Psychoneuroendocrinology 37, 1277–1285. 99, 69–78. Schwartz, C.E., et al., 1990. Maternal expressed emotion and Tu, M.T., Lupien, S.J., Walker, C.D., 2006. Diurnal salivary cortisol parental affective disorder: risk for childhood depressive disorder, substance abuse, or conduct disorder. J. Psychiatr. levels in postpartum mothers as a function of infant feeding Res 24, 231–250. choice and parity. Psychoneuroendocrinology 31, Shahrokh, D., et al., 2010. Oxytocin-dopamine interactions 812–824. mediate variations in maternal behavior in the rat. Turner, K.M., et al., 2008. Women’s views and experiences of Endocrinology 151, 2276–2286. antidepressants as a treatment for postnatal depression: a Sharp, D.J., et al., 2010. A pragmatic randomised controlled trial to qualitative study. Family Pract 25, 450–455. compare antidepressants with a community-based van Ijzendoorn, M.H., et al., 2011. The impact of oxytocin psychosocial intervention for the treatment of women with administration on charitable donating is moderated by postnatal depression: the RESPOND trial. Health Technol. experiences of parental love-withdrawal. Front. Psychol 2, 258. Assess 14, iii–iv, ix–xi, 1–153. Van IJzendoorn, M.H., Bakermans-Kranenburg, M.J., 2012. A sniff Sit, D.K., Wisner, K.L., 2009. Identification of postpartum of trust: meta-analysis of the effects of intranasal oxytocin depression. Clin. Obstetr. Gynecol 52, 456–468. administration on face recognition, trust to in-group, Skalkidou, A., et al., 2010. Biological aspects of postpartum and trust to out-group. Psychoneuroendocrinology 37, depression. Womens’ Health 8, 659–672. 438–443. Skrundz, M., et al., 2011. Plasma oxytocin concentration during van Leengoed, E., Kerker, E., Swanson, H.H., 1987. Inhibition of pregnancy is associated with development of postpartum post-partum maternal behaviour in the rat by injecting an depression. Neuropsychopharmacology 36, 1886–1893. oxytocin antagonist into the cerebral ventricles. J. Endocrinol Slattery, D.A., Neumann, I.D., 2010a. Chronic icv oxytocin 112, 275–282. attenuates the pathological high anxiety state of selectively van Londen, L., et al., 1997. Plasma levels of arginine vasopressin bred Wistar rats. Neuropharmacology 58, 56–61. elevated in patients with major depression. Slattery, D.A., Neumann, I.D., 2010b. Oxytocin and major Neuropsychopharmacology 17, 284–292. depressive disorder: experimental and clinical evidence for Wang, L., et al., 2011. Prevalence and risk factors of maternal links to aetiology and possible treatment. Pharmaceuticals 3, depression during the first three years of child rearing. 702–724. J. Womens Health (Larchmt) 20, 711–718. Sockol, L.E., Epperson, C.N., Barber, J.P., 2011. A meta-analysis of Watkins, S., et al., 2011. Early breastfeeding experiences treatments for perinatal depression. Clin. Psychol. Rev. 31, and postpartum depression. Obstetr. Gynecol 118, 839–849. 214–221. Sroufe, L.A., et al., 2005. The Development of the Person. Guilford Weikum, W.M., et al., 2013. The impact of prenatal serotonin Press, New York, NY. reuptake inhibitor (SRI) antidepressant exposure and Stern, D.N., 1995. The Motherhood Constellation: A Unified View maternal mood on mother–infant interactions at 3 months of of Parent–Infant Psychotherapy. Basic Books, New York, NY. age. Infant Behav. Dev 36, 485–493. Strathearn, L., et al., 2009. Adult attachment predicts Weinberg, M.K., Tronick, E.Z., 1998. The impact of maternal maternal brain and oxytocin response to infant cues. psychiatric illness on infant development. J. Clin. Psychiatry Neuropsychopharmacology 34, 2655–2666. Strathearn, L., 2011. Maternal neglect: oxytocin, dopamine and 59, 53–61. the neurobiology of attachment. J. Neuroendocrinol. 23, Wiklund, I., Mohlkert, P., Edman, G., 2010. Evaluation of a brief 1054–1065. cognitive intervention in patients with signs of postnatal Strathearn, L., et al., 2012. Maternal oxytocin response during depression: a randomized controlled trial. Acta Obstet. mother–infant interaction: associations with adult Gynecol. Scand 89, 1100–1104. temperament. Horm. Behav 61, 429–435. Windle, R.J., et al., 2004. Oxytocin attenuates stress-induced c-fos Stuebe, A.M., et al., 2012. Failed lactation and perinatal mRNA expression in specific forebrain regions associated with depression: common problems with shared neuroendocrine modulation of hypothalamo–pituitary–adrenal activity. mechanisms?. J. Womens Health (Larchmt) 21, 264–272. J. Neurosci 24, 2974–2982. Stuebe, A.M., Grewen, K., Meltzer-Brody, S., 2013. Association Winslow, J.T., et al., 2003. Rearing effects on cerebrospinal fluid between maternal mood and oxytocin response to oxytocin concentration and social buffering in rhesus breastfeeding. J. Womens Health (Larchmt) 22, 352–361. monkeys. Neuropsychopharmacology 28, 910–918. 232 brain research 1580 (2014) 219–232

Wisner, K.L., Wheeler, S.B., 1994. Prevention of recurrent Yim, I.S., et al., 2009. Risk of postpartum depressive symptoms postpartum major depression. Hosp. Commun. Psychiatry 45, with elevated corticotropin-releasing hormone in human 1191–1196. pregnancy. Arch. Gen. Psychiatry 66, 162–169. Wisner, K.L., et al., 2006. Postpartum depression: a randomized Yonkers, K.A., et al., 2008. Pharmacologic treatment of trial of sertraline versus nortriptyline. J. Clin. postpartum women with new-onset major depressive Psychopharmacol 26, 353–360. disorder: a randomized controlled trial with paroxetine. Wisner, K.L., Moses-Kolko, E.L., Sit, D.K., 2010. Postpartum J. Clin. Psychiatry 69, 659–665. depression: a disorder in search of a definition. Arch. Womens Yonkers, K.A., et al., 2009. The management of depression during Ment. Health 13, 37–40. pregnancy: a report from the American Psychiatric Workman, J.L., Barha, C.K., Galea, L.A., 2012. Endocrine substrates Association and the American College of Obstetricians and of cognitive and affective changes during pregnancy and Gynecologists. Gen. Hospital Psychiatry 31, 403–413. postpartum. Behav. Neurosci. 126, 54–72. World Health Organization, 1992. The ICD-10 Classification of Yoshida, M., et al., 2009. Evidence that oxytocin exerts anxiolytic Mental and Behavioural Disorders: Clinical Descriptions effects via oxytocin receptor expressed in serotonergic and Diagnostic Guidelines. World Health Organization, neurons in mice. J. Neurosci 29, 2259–2271. Geneva. Zetzsche, T., et al., 1996. Nocturnal oxytocin secretion is reduced Yatzkar, U., Klein, E., 2009. Intranasal oxytocin in patients with in major depression. Biol. Psychiatry 39, 584. post traumatic stress disorder: a single dose pilot double blind Zonana, J., Gorman, J.M., 2005. The neurobiology of postpartum crossover study. Clin. Neuropsychopharmacol 32, S84. depression. CNS Spectr 10, 792–799, 805. brain research 1580 (2014) 57–68

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Oxytocin and social cognition in rhesus macaques: Implications for understanding and treating human psychopathology

Steve W.C. Changa,b,c,n, Michael L. Plattb,c,d,e,f aDepartment of Psychology, Yale University, New Haven, CT 06511, USA bDuke Institute for Brain Sciences, Duke University, Durham, NC 27708, USA cCenter for Cognitive Neuroscience, B203 Levine Science Research Center, Duke University, Box 90999, Durham, NC 27708, USA dDepartment of Neurobiology, Duke University School of Medicine, Durham, NC 27710, USA eDepartment of Psychology and Neurosciences, Duke University, Durham, NC 27708, USA fDepartment of Evolutionary Anthropology, Duke University, Durham, NC 27708, USA article info abstract

Article history: Converging evidence from humans and non-human animals indicates that the neuro- Accepted 4 November 2013 hypophysial hormone oxytocin (OT) evolvedtoserveaspecializedfunctioninsocial Available online 11 November 2013 behavior in mammals. Although OT-based therapies are currently being evaluated as fi Kyewords: remedies for social de cits in neuropsychiatric disorders, precisely how OT regulates Oxytocin complex social processes remains largely unknown. Here we describe how a non- Social cognition human primate model can be used to understand the mechanisms by which OT Rhesus macaques regulates social cognition and thereby inform its clinical application in humans. We Amygdala focus primarily on recent advances in our understanding of OT-mediated social Social decision-making cognition in rhesus macaques (Macaca mulatta), supplemented by discussion of recent Social vigilance work in humans, other primates, and rodents. Together, these studies endorse the Intranasal hypothesis that OT promotes social exploration both by amplifying social motivation Nebulizer and by attenuating social vigilance. Inhalation This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2013 Elsevier B.V. All rights reserved.

1. Introduction ancestral functions of OT have been co-opted to serve social functions, such as promoting maternal behavior (Champagne Oxytocin (OT) is an evolutionarily conserved nonapeptide et al., 2001; Pedersen et al., 1982), fostering pair-bonding and that mediates female sexual intercourse, parturition, lacta- affiliative behaviors (Cushing and Carter, 2000; Smith et al., tion, as well as water regulation, and anxiolytic functions 2010; Snowdon et al., 2010; Young and Wang, 2004), encoura- (Donaldson and Young, 2008). In highly social animals, these ging in-group bias (Dreu et al., 2010), reducing social vigilance

nCorresponding author at: Department of Psychology, Yale University, New Haven, CT 06511, USA. E-mail addresses: [email protected], [email protected] (S.W.C. Chang).

0006-8993/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.brainres.2013.11.006 58 brain research 1580 (2014) 57–68

(Ebitz et al., 2013; Heinrichs et al., 2003) and amplifying other- (Watson and Platt, 2012). Although there are undoubtedly some regarding behaviors (Barraza et al., 2011; Chang et al., 2012; differences between humans and rhesus monkeys (Byrne and Van IJzendoorn et al., 2011). Although there appears to be a Whiten, 1988), such as the strength of prosociality, rhesus large range of OT-mediated effects, one might argue that macaques are outstanding models for studying the neural some functions of OT may be common to most, if not all, of mechanisms underlying psychiatric disorders marked by social OT-mediated social cognition. The anxiolytic, approach-pro- deficits. Due to their remarkably similar anatomy and physiol- moting, and tolerance-enhancing roles of OT (Amico et al., ogy to humans, rhesus macaques have long served as the gold 2004; Averbeck, 2010; Heinrichs et al., 2003; Kemp and standard for electrophysiological, pharmacological, and lesion- Guastella, 2010; Neumann et al., 2000a; Riem et al., 2011; Ring based investigations into complex cognitive processes. In this et al., 2006; Uvnäs-Moberg et al., 1994; Waldherr and Neumann, review we highlight recent advances in understanding how OT 2007; Yoshida et al., 2009; Young, 2002)mayserveasfounda- influences social behavior in rhesus macaques, paving the way tional substrates that promote social exploration and interac- for future investigations into the neural mechanisms mediating tion while, typically, suppressing social avoidance. these influences. A large number of studies have been conducted to probe the role of OT in regulating social behavior in both healthy and pathological states (Bartz et al., 2011; De Dreu, 2012; 2. Inhaling OT increases its concentration Guastella et al., 2012; Heinrichs and Domes, 2008; Insel, 2010; in the brain MacDonald and Feifel, 2013; Meyer-Lindenberg et al., 2011). Nevertheless, the neural mechanisms through which OT Numerous human studies have demonstrated that intrana- regulates social behavior and cognition—particularly in sally administered OT can modulate complex social cogni- humans—remain poorly understood. Standard noninvasive tion. One of the most exciting findings from recent OT studies neuroscientific techniques, such as functional magnetic reso- is that the peptide appears to rescue some social deficits in nance imaging (fMRI) and transcranial magnetic stimulation individuals with psychopathological conditions (for a review, (TMS), used to study human brain and cognition are limited see: Insel, 2010; Meyer-Lindenberg et al., 2011). The clinical in their capacity to reveal the neuronal and circuit mechan- and basic science communities are currently working isms that mediate the regulation of social behavior and together to translate basic OT research into useful and safe cognition by OT. Conversely, rodent models permit exqui- OT therapies for social disorders (Miller, 2013). Nevertheless, sitely fine dissection of these neural pathways but lack the whether or not intranasal administration of OT actually behavioral complexity of human social function. translocates the peptide into the central nervous system Compared to other animals, primates, including humans, (CNS) remained unknown until recently. In humans, the are unique in that they show remarkably complex social closest demonstration was for arginine vasopressin (AVP), behavior in a society typically made up of many individuals. another neurohypophysial hormone with social functions Adapting to increasing social complexity may have played a closely related to OT and differing in only two amino acids. major role in primate brain evolution (Dunbar and Shultz, 2007). Intranasally administered AVP effectively increases CNS AVP For example, across primate species, social complexity, as concentration for a long duration (480 min.) in a dose- measured by group size, strongly predicts forebrain volume dependent manner (Born et al., 2002). More recently, data (after correcting for body size) (Dunbar, 1998). Although rodents from rhesus macaques demonstrated that aerosolized OT offer the opportunity for exploitation of powerful molecular using a nebulizer system (Fig. 1A) effectively reaches the genetic techniques, their social behavior is not very similar to brain. Using a pediatric nebulizer, we recently showed that the social behavior of humans. While molecular genetic tech- inhaled oronasal administration of OT increases its concen- niques are only beginning to be developed for use in non- tration in the cerebrospinal fluid (CSF), measured at 30 min human primates (Diester et al., 2011; Sasaki et al., 2009), their post-delivery (Chang et al., 2012)(Fig. 1B). A recent micro- social behavior is much more similar to the social behavior of dialysis study in rats and mice has demonstrated that nasal humans. administration of OT increases levels in the central nervous Here we argue that a rhesus macaque model (Macaca system (sampled from the amygdala and hippocampus), mulatta)caneffectivelybridgethisgap.Rhesusmacaquesare peaking 30–60 min from the time of nasal delivery Old World monkeys that live in large, hierarchical, and mixed- (Neumann et al., 2013). Subsequent work in primates from sex social groups, that last shared a common ancestor with another laboratory reported that the inhaled administration humans some 25 million years ago (Smuts, 1987). Critically, of aerosolized OT effectively elevates OT levels in the CSF in rhesus macaques display basic aspects of complex social rhesus macaques, but the application of intranasal spray, behaviors that are typically considered ‘uniquely human’ which has been the standard method in studies in humans, (Frith and Frith, 2007; Saxe, 2006). These include social imitation does not (Modi et al., in press). (Ferrari et al., 2006; Subiaul, 2004), prosocial behaviors (Chang Anecdotally, monkeys readily accept and tolerate nebuli- et al., 2011; Masserman et al., 1964), as well as understanding zation (e.g., training takes less than a week), a technique that others' perceptions (Flombaum and Santos, 2005; Santos is routinely used in babies and young children to administer et al., 2006). Mental capacities like these might be fundamental drugs like albuterol to alleviate breathing difficulties, suggest- building blocks for empathy and theory of mind. Such simila- ing that this method may prove both effective and acceptable rities in social behaviors make rhesus macaques excellent in young patients with neuropsychiatric disorders. This models for studying neuropsychiatric conditions accompanied tolerability is particularly desired for an early OT intervention by complex social deficits, such as autism spectrum disorders in young children or even infants. Further research will be brain research 1580 (2014) 57–68 59

OT treatment (2 monkeys) Saline Training 60 treatment (4 monkeys)

~ 1 week of 50 accustomization

40 * 30

confirm

in the CSF (pg/ml) level OT CSF 20 Inhalation treatment 10

100 * * 1

aggression -20

32 Frequency of * * 18 24 36 * * 0.5 * 20 Duration of Baseline CSF OT Level (pg/ml) Level OT affiliative behviors (sec)

8 0 18 24 36 18 24 36 Ages of the monkeys Ages of the monkeys Mother-reared Nursery-reared

Fig. 1 – OT concentration in the cerebrospinal fluid (CSF) and social development in rhesus macaques. (A) A nebulizer-based inhalation setup using a nebulizer (PARI Baby Nebulizer) in rhesus macaques. Notice that both nose and mouth are covered with the inhalation mask. Top, a training phase in which juice rewards aredeliveredthroughthenebulizermask(aboutaweek).Bottom,a treatment phase in which aerosolized OT or saline solutions are delivered to the monkeys. (B) OT concentration in the CSF after inhaling 25 IU of OT (red) or saline (dark gray) in monkeys (5 min of nebulization at a 5 IU/min). The CSF samples were obtained from cervical punctures at 30 min post-delivery. Colored outlines on data points identify individual monkeys. Arrows with matching n colors emphasize the CSF OT concentrations following saline (baseline) and OT inhalations within the same two monkeys. , Po0.05, Welch two-sample t test. (C) Different levels of baseline CSF OT for mother-reared and nursery-reared male rhesus macaques across 18, 24, and 36 months of age. n, Po0.05, ANOVA. (D). Enhanced duration of affiliative behavioral engagements (e.g., allogrooming and reciprocal male mounting) in mother-reared compared to nursery-reared monkeys. The inset shows reduced frequency of aggressive behaviors toward conspecifics for mother-reared compared to nursery-reared monkeys. n, Po0.05, ANOVA. [B adapted from: Chang, S.W.C., Barter, J.W., Ebitz, R.B., Watson, K.K., Platt, M.L. (2012).Inhaledoxytocinamplifies both vicarious reinforcement and self reinforcement in rhesus macaques (Macaca mulatta). Proc. Natl. Acad. Sci. 109, 959–964, Copyright (2012) National Academy of Sciences, U.S.A.; C, D adapted from: Winslow, J.T., Noble, P.L., Lyons, C.K., Sterk, S.M., Insel, T.R. (2003).Rearing effects on cerebrospinal fluid oxytocin concentration and social buffering in rhesus monkeys. Neuropsychopharmacology 28, 910–918, Copyright (2003) Nature Publishing Group]. 60 brain research 1580 (2014) 57–68

needed to confirm whether the enhanced efficacy of oronasal since some variables other than early developmental style nebulization of OT translates to humans. The outcome of could be driving the changes in either central or peripheral OT such investigations will have a critical impact on how OT is levels. Furthermore, it is likely that the hypothalamic–pitui- administered to individuals in clinical settings as well as how tary–adrenal (HPA) system may contribute to the effects of peripheral versus central effects of OT on social cognition are different rearing conditions, resulting in changes in both social understood. behavior and OT function. Nevertheless, OT levels (present Although effective, non-invasive, and easy to administer, either in the central or periphery system) are correlated with it is virtually impossible to determine the precise amount long-lasting social behaviors shaped from the very first encoun- delivered to the brain using intranasal or inhaled adminis- ters between infants and mothers. These findings strongly tration. During the intranasal or inhalation process, some endorse the initiation of OT-based therapies as early as possible amount of OT is bound to be absorbed peripherally or simply (using a tolerable nebulizer method) after detection of a leak out. This limitation of intranasal and inhalation delivery neuropsychiatric disorder with social deficits, such as autism. may limit the attractiveness of OT therapy if it requires a precise dose to be effective and safe. Future studies in animals should focus on estimating, as accurately and 4. OT amplifies intrinsic social motivation reliably as possible, the amount of OT delivered versus not in rhesus macaques delivered to the brain following intranasal or inhalation administration. Much of our social interactions are driven by reinforcement, both direct rewards we receive ourselves and the reward we experience when good things happen to others (i.e., vicarious 3. OT is critical for normal social development reinforcement (Bandura et al., 1963; Berber, 1962)). How might in rhesus macaques OT in the CNS shape both self and vicarious reinforcement during social decisions? We recently developed a social Like humans (Hinde, 1974), the development of social beha- reward-allocation paradigm in rhesus macaques (Chang vior in infant rhesus monkeys depends critically on how they et al., 2011), in which an actor monkey makes a series of are raised. Infant monkeys raised away from their mothers decisions to deliver juice rewards to himself, to a recipient later show a number of social deficits and stress-related monkey present in the room, to both simultaneously, or to no abnormalities (Champoux et al., 1992; Harlow et al., 1965; one. When given a choice between rewarding self and Sackett, 1984; Suomi, 1991; Winslow et al., 2003). Based on the rewarding self and the other monkey together at no addi- importance of OT in mother-infant relationships and tional cost, actor monkeys prefer to deliver juice rewards to bonding-related behaviors, such as mutual gaze, vocaliza- themselves only, displaying an antisocial preference (Chang tions, and affiliative touching (Feldman et al., 2007; Francis et al., 2011). Moreover, not surprisingly, when given a choice et al., 2000; Galbally et al., 2011; Pedersen, 1997; Riem et al., between rewarding themselves only and the recipients only, 2011; Strathearn et al., 2009), it seems likely that the peptide the actor monkeys prefer to deliver juice almost exclusively plays a crucial role in the early development of social to themselves (Chang et al., 2012, 2013a). On the other hand, behavior. Indeed, OT seems to shape the social behavior of when choosing between rewarding the other monkey and both mothers and their infants. Evidence suggests that rewarding no one, the same actor monkeys prefer to deliver individual differences in maternal behaviors (e.g., attachment juice rewards to the recipient monkey, demonstrating an styles) are linked to oxytocinergic systems in reward-related other-regarding preference that in humans would be consid- brain regions, such as ventral striatum and amygdala (Francis ered prosocial (Chang et al., 2011, 2012, 2013a). et al., 2000; Strathearn et al., 2009). In rhesus macaques, Oronasal nebulization of OT amplifies these self-regarding baseline CSF OT levels, but not CSF AVP levels, are signifi- and other-regarding preferences in rhesus macaques (Chang cantly reduced in males raised in a nursery compared to et al., 2012). When choosing between allocating juice rewards peers raised by their mothers (Winslow et al., 2003)(Fig. 1C), to a recipient monkey and no one at all, exogenous OT confirming a specialized role of OT in early social develop- enhances the baseline prosocial bias (Fig. 2A). Amplified ment. Furthermore, mother-reared monkeys display substan- prosocial preference is accompanied by an increase in gaze tially more affiliative behaviors (e.g., allogrooming and shifts to the recipient monkeys' face region following OT reciprocal male mounting behaviors) (Fig. 1D) and less (Fig. 2B). Critically, the OT-treated actors do not increase such aggressive behaviors (Fig. 1D inset) toward other individuals, gaze shifts to the recipient monkeys when the juice rewards are (For interpretation of the references to color in this figure delivered to the recipients by the experimenter (i.e., cued legend, the reader is referred to the web version of this article.) condition) (Chang et al., 2012). These OT-mediated gaze pat- suggesting a link between CNS OT levels and affiliative social terns during active decision-making suggest a link between tendencies developed later in their lives (18–36 months of age) reinforcing action and observation of the rewarding experience (Winslow et al., 2003). Similarly, human children who are of the recipient monkeys. Such OT-mediated increases in neglected by their parents immediately after birth also show attention to social stimuli like eyes and faces have been significantly reduced urinary OT levels later on (on average age well documented in healthy humans as well as those with of 4.5 years old) (Wismer-Fries et al., 2005). It remains to be autism spectrum disorders (e.g., Andari et al., 2010; Guastella determined whether urinary OT levels are directly associated et al., 2008). with behaviors mediated by the central oxytonergic system. Furthermore, inhaled OT significantly increases reaction Overall, one needs to be cautious in interpreting these results times when the actor monkeys choose between donating juice brain research 1580 (2014) 57–68 61

Estimated Time (min) OT 04590135180 Saline

1.0 other vs. neither 25

* bias toward 20 * other over neither 0.5 other over self neither over self 15 Number of Social Gaze

0.0 10 0510152025 Trial Number (x100) Preference Index OT Saline other vs. neither bias toward -0.5 self vs. other self over other self over neither OT self vs. neither neither over other Saline * 180

Un-jittered self vs. other 160 -1.0 * -0.96 140

0510152025 -0.98 120

Trial Number (x100) -1.00

Reaction Time (ms) Time Reaction 100 05101520 self self other vs. vs. vs. neither other neither Decision context

Fig. 2 – OT promotes social motivation in rhesus macaques. (A) Choice preference index following inhaled OT (red) or saline (gray) administration for rewards delivered to: other (recipient monkey) versus (vs.) neither, self (actor) vs. other, and self vs. neither in the social reward-allocation task. Data points from self vs. other and self vs. neither are jittered on the left plot for visibility. The inset shows unjittered data from self vs. other trials. n, Po0.05, Welch two-sample t test. (B) Number of gaze shifts made to the recipient monkey after reward delivery over the course of each session on other vs. neither trials. n, Po0.05, Welch two-sample t test. (C) OT selectively increases the decision deliberation time in the other vs. neither context (choosing to deliver juice rewards to other or no one) in which actor rhesus monkeys show a preference for delivering juice rewards to another monkey. n, Po0.05, Welch two-sample t test. [A–C adapted from: Chang, S.W.C., Barter, J.W., Ebitz, R.B., Watson, K.K., Platt, M.L. (2012). Inhaled oxytocin amplifies both vicarious reinforcement and self reinforcement in rhesus macaques (Macaca mulatta). Proc. Natl. Acad. Sci. 109, 959–964, Copyright (2012) National Academy of Sciences, U.S.A.].

rewards to the recipient monkey and no one (Chang et al., 2012) social preferences in rhesus macaques is consistent with (Fig. 2C), inviting the possibility that OT promotes prosocial the effects of OT manipulation on prosocial choices in pair- choices by increasing internal deliberative processing in rhesus bonding marmosets (Callithrix penicillata). In that study, treat- macaques. Such increased deliberation processes might be ment with an OT receptor antagonist effectively eliminated necessary for enhancing prosocial behaviors in highly despotic species-typical food sharing behavior between paired male rhesus macaques, compared to humans who seem to sponta- and female marmosets (Smith et al., 2010). Taken together, neously prefer, and thus show faster reactions times for, these observations are consistent with the hypothesis that prosocial decisions (Rand et al., 2012). These results suggest OT regulates the gain of pre-existing social preferences rather that OT enhances vicariously reinforcing actions by possibly than changing their fundamental character. coupling reinforcement and social observation. By contrast, when choosing between delivering juice rewards to themselves and to the recipients, inhaled OT 5. OT relaxes social vigilance, thereby amplifies the self-regarding preference (i.e., delivering juice permitting social exploration in rhesus macaques to only themselves over only the recipients), essentially eliminating the small number of prosocial choices in this One way to promote social interactions is by modulating the competitive context (Fig. 2A). Therefore, as in humans (Bartz social state of an animal in order to encourage social et al., 2011), OT seems to elicit context-specific social beha- exploration. To investigate the role of OT in modulating the viors in rhesus macaques. OT-mediated enhancement of social state in rhesus macaques, our group has recently yterdcini oky'seistpclcocst iwdmnn aeiae A,O eetvl lmntstypical eliminates selectively OT faces. (A), monkey images dominant face view dominant to view choosing of to facilitation choices time species-typical reaction monkeys' in reduction the by ongtv ailepesos scoeredciooy3,1748 38, Psychoneuroendocrinology expressions. facial negative to from: agt ih,O eetvl eue h yia acd nefrne maue ytemgiueo de of magnitude the by (measured to interferences relative location saccade neutral typical a the at reduces presented selectively (lower) distractors OT nonsocial Right, and target. (upper) a distractors social with trials from residuals distractors. aeoisatrihldO rpaeo ih,atninlba crsfrdrc essaetdgz rasatrihldO or stimuli OT facial to inhaled attention after enhancing trials of gaze trend a averted shows [A versus it (right). direct whereas (left), gaze for stimuli direct scores face with bias emotional to attentional attention Right, reduces placebo. OT placebo. or OT inhaled after categories hssmcqe rc al cd c.10 11630 110, Sci. Acad. Natl. Proc. macaque. rhesus ove oiatfc mgsi aki hc oky hoet iwteiae fdmnn ihsau monkeys, status high dominant of images the view to ( choose square monkeys gray or which perinea, in female task monkeys, a status in low subordinate images face dominant view to 3 Fig. 62 – ar .. oi . ibr,E,Yug ..(2013) L.J. Young, E., Siebert, M., Modi, L.A., Parr, Trgltssca iiac nreu aaus A nae Tslcieyrdcsmnes pce-yia choices species-typical monkeys' reduces selectively OT Inhaled (A) macaques. rhesus in vigilance social regulates OT n , P

o Attentional bias score nonsocial distractor social distractor Fraction image choices .1 ofroicretdps o ue S.()Lf,Atninlba crsfrfca xrsinstimuli expression facial for scores bias Attentional Left, (D) LSD. Tukey doc post Bonferroni-corrected 0.01, Facial expressionvs.nonsocialimages l xrsintrasbared-teeth threats expression all saccade – dpe from: adapted C * Saccade deflection

(sum of residuals, degrees) 57–68 (2014) 1580 research brain 106 134 bt,RB,Wto,KK,Pat ..(2013) M.L. Platt, K.K., Watson, R.B., Ebitz, 50 78 – 13,Cprgt(03 ainlAaeyo cecs ... adapted D U.S.A.; Sciences, of Academy National (2013) Copyright 11635, aielowOT saline OT saline * nrnsloyoi eetvl teutsreu oky'attention monkeys' rhesus attenuates selectively oxytocin Intranasal . OT placebo (10 IU)

enre l,2005 al., et Deaner Choice – 76 oyih 21)Elsevier]. (2013) Copyright 1756, reaction time (ms) n , P o nonsocial distractor social distractor high OT (25 IU)

.5 othcLD C et xmlso saccade of Examples Left, (C) LSD. hoc post 0.05, Attentional bias score Direct vs. averted gaze Images chosentoview ). xtcnbut oilvglnei the in vigilance social blunts Oxytocin . n , P * o .5 ue S.()Accompanied (B) LSD. Tukey 0.05, saline OT fl cin)t social to ections) brain research 1580 (2014) 57–68 63

investigated social vigilance behavior in male monkeys permissive, rather than promotional, role in social behavior following OT inhalation (Ebitz et al., 2013). When monkeys in rhesus macaques, and perhaps humans as well. choose whether to acquire different types of visual informa- Increasing social exploration should have obvious conse- tion about the local social context (viewing different social quences for forming and maintaining social bonds. Research images), OT selectively reduces species-typical tendencies to in non-human primates as well as in humans (Bartz et al., view the faces of dominant monkeys, a threatening but 2011; De Dreu, 2012; Guastella et al., 2012; Heinrichs and highly informative stimulus (Fig. 3A). Inhaled OT also elim- Domes, 2008; Insel, 2010; MacDonald and Feifel, 2013; Meyer- inates the privileged processing for dominant faces over Lindenberg et al., 2011) suggests that OT might be an other images, effectively slowing monkeys down when mak- important component of the neuroendocrinological regula- ing this particular decision (Fig. 3B). Moreover, OT substan- tion of social relationships. In non-human animals, social tially attenuates species-typical distraction by the peripheral relationships can be studied by examining naturally- flash of images of unfamiliar monkey faces, indexed by a occurring social bonds (Brent et al., in press, 2013). A recent reduction in gaze deflection towards them (Fig. 3C). These study in wild chimpanzees (Pan troglodytes) measured urinary findings endorse the idea that OT helps regulate species- OT levels after grooming bouts with different partners typical social vigilance. Reducing social vigilance state in turn (Crockford et al., 2013). urinary OT levels after grooming could free up cognitive resources and promote social explora- behaviors predicted the strength of social bonds among the tion (Ebitz et al., 2013). partners, and, surprisingly, this effect was not explained by Consistent with the role of OT in reducing social vigilance genetic relatedness. Therefore, the action of OT in social state, another recent study in rhesus macaques reported that processing seems to critically depend on the history of social inhaled OT selectively reduces attention to emotional facial interactions between two individuals. Future studies exam- expressions while enhancing attention to faces with direct ining how prior social experience influences neural circuits gaze (Parr et al., 2013)(Fig. 3D), which is a threatening gesture regulated by OT should reveal important new insights into in macaques. Moreover, OT delivered intranasally to squirrel the neural basis of social relationships. monkeys (Suimiri sciureus) attenuates stress responses by lowering adrenocorticotropic hormone (ACTH) (i.e., cortico- tropin) levels following 90 min of social isolation (Parker et al., 6. OT receptor distribution in the brain 2005). Such OT-mediated reduction in ACTH levels suggests that OT regulates social stress by acting through the hypotha- Current knowledge regarding the distribution of OT receptors lamic–pituitary–adrenal (HPA) axis. Taken together, these in the mammalian brain is derived primarily from rodents, findings suggest that OT may facilitate social interactions though there have been some studies in humans and mon- by lowering social vigilance and reducing social stress (Carter, keys. In rats, some of the regions with high densities of OT 1998; Chang et al., 2013b; Ebitz et al., 2013; Neumann et al., receptor include the olfactory nucleus, hypothalamic regions, 2000b; Uvnäs-Moberg et al., 1994). and the central amygdala (Tribollet and Barberis, 1996; Following intranasal OT (Syntocinon spray, Novartis), Tribollet et al., 1992). Importantly, varying levels of OT humans spend more time visually inspecting the eye region receptors were detected across many functional subsystems, of faces compared to placebo (i.e., all inactive ingredients such as the limbic, basal ganglia, cortical, brainstem and the except for the peptide) (Gamer et al., 2010; Guastella et al., spinal cord (Tribollet et al., 1992). Autoradiographic studies 2008). Moreover, intranasal OT improves test scores for the have demonstrated that the distributions of OT receptors Reading the Mind in the Eyes Test (RMET) (Domes et al., differ greatly between monogamous prairie voles (Microtus 2007), which requires participants to infer mental states ochrogaster) and polygamous montane voles (Microtus monta- from photographs of eyes (Baron-Cohen et al., 2001). Intra- nus), in a manner consistent with differences in social nasal OT also increases attention to the eyes in individuals bonding between these species (Insel and Shapiro, 1992; with social deficits. Individuals with autism, for example, Young and Wang, 2004). Compared to montane voles, OT spend more time attending to the eye region of faces receptors in prairie voles are densely distributed in prelimbic following intranasal OT compared to placebo (Andari areas, including a region analogous to the primate anterior et al., 2010). These results in humans may at first appear cingulate gyrus, the nucleus accumbens, the bed nucleus of contradictory to the social vigilance results from rhesus stria terminalis, and the lateral amygdala (Insel and Shapiro, monkeys (Ebitz et al., 2013; Parr et al., 2013). Although 1992; Young et al., 1996). In human and non-human primates, exogenous OT reduces social vigilance to social threats in immunoreactive studies have yielded mixed results with rhesus macaques (Fig. 3A), it increases the amount of time respect to finding OT receptors in the same areas as in rhesus spend looking at faces and eyes, as in humans (see rodents (Caffé et al., 1989; Jenkins et al., 1984; Sofroniew Discussion in Ebitz et al., 2013). Moreover, when deciding et al., 1981; Wang et al., 1997). Similar to findings in rats whether or not to allocate juice rewards to another monkey (Audigier and Barberis, 1985; Tribollet et al., 1992), OT immu- present in the room, inhaled OT increases gaze to the face noreactive cells were found in the hypothalamus, the bed of the recipient monkey (Fig. 2B) (Chang et al., 2012). Based nucleus of the stria terminalis, and the medial amygdala in on these results, we conjecture that reduced vigilance common marmosets (Callithrix jacchus)(Wang et al., 1997), the might play a permissive role in social approach behavior, new world primates that pair-bond similar to prairie voles. favoring increased attention to faces and eyes and subse- Nevertheless, studies in human and non-human primates quent enhancements in social cognition. Further studies have also found some notable differences in OT receptor will be necessary to test the idea that OT plays a distributions compared to rodents (Boccia et al., 2013; Caffé 64 brain research 1580 (2014) 57–68

et al., 1989; Jenkins et al., 1984; Loup et al., 1991). A recent fear responses through a specialized circuit via the central immunohistochemical study in postmortem human brains amygdala (Huber, 2005; Viviani et al., 2011). This OT-mediated localized OT receptors in the central and basolateral amyg- autonomic circuit might be also utilized for regulating social dala, medial preoptic area, hypothalamus, anterior cingulate vigilance. Whether there are parallel oxytocinergic circuits for cortex, and ventrolateral septum, among others (Boccia et al., social motivational processing in the amygdala (e.g., in the 2013). In contrast to rat brains, the authors did not localize OT basolateral subdivisions) remains unknown. Taken together, receptors in the hippocampus, supraoptic nucleus, and these findings suggest that OT may promote social exploration nucleus accumbens, among others, in their human brain through modulations in socioemotional processing within the tissues (Boccia et al., 2013). Clearly, resolving the distribution amygdala and its interconnected limbic circuitry. and binding affinities of OT receptors in human and non- Interestingly, one study on human participants found that human primate brains is a high priority and will require the BOLD signals in the amygdala in women, unlike in men, development of new ligands for OT receptors. Despite differ- actually increase when viewing fearful faces (Domes et al., ences between species and these methodological limitations, 2010). One possible mechanism for this gender-specific effect it is interesting to note that some regions appear to express of OT in amygdala activations is the interactions between OT receptors across species, notably the amygdala. In the oxytocin receptors and gonadal steroid hormones (Domes next section, we review current understanding of the role of et al., 2010). It has been shown that estradiol promotes the amygdala in OT-mediated social cognition. oxytocin receptor binding by increasing the oxytocin receptor concentrations, whereas progesterone inhibits oxytocin receptor by repressing its synthesis (Choleris et al., 2008; 7. Amygdala and context-specific Gimpl et al., 2002; Nissenson et al., 1978). Interactions OT-mediated modulations between OT and gonadal steroids have also been reported in monkeys as well. In male squirrel monkeys, ventricular The amygdala, which possesses an unusually high density of injection of OT enhances sexual and aggressive behaviors OT receptors (Francis et al., 2000; Insel and Shapiro, 1992; toward a familiar female monkey in dominant males who Young et al., 1996), may be a key site through which OT have high testosterone levels, but enhances affiliative beha- influences social behavior. The specialized functions of the viors like touching and huddling in subordinate males who amygdala in social cognition has been extensively reported in have low testosterone levels (Winslow and Insel, 1991). Such healthy humans and amygdala-damaged human patients (for gender-specific effects of OT have been documented numer- a review, Adolphs, 2010). Experimental lesions to the amyg- ously in prairie voles (Microtus ochrogaster), and are verified to dala in rhesus macaques lead to stereotyped social deficits, be centrally-mediated based on ventricular peptide manip- such as reduced aggression and increased submission ulation experiments. For instance, the context-specific role of (Amaral, 2002; Dicks et al., 1969; Kling, 1972; Kluver and OT in pair-bonding and mate-guarding aggression is more Bucy, 1939; Meunier et al., 1999; Rosvold et al., 1954). Further- critical to female voles, whereas the role of AVP in these two more, neonatal bilateral amygdala lesions in rhesus maca- types of behaviors is more critical to male voles (Bales and ques result in blunted affective expressions in adulthood Carter, 2003; Cushing and Carter, 2000; Insel and Hulihan, toward both positive and negative stimuli (Bliss-Moreau et al., 1995; Winslow et al., 1993; Young and Wang, 2004). Further 2011). Moreover, neurons in the primate amygdala signal research is required to understand how and where in the motivational values and is causally implicated in reward- brain OT system interacts with sex hormones to differentially guided learning (Baxter and Murray, 2002; Paton et al., 2006). influence social cognition in men and women. These inter- These findings place the amygdala as an ideal anatomical actions may govern the type of social states that is encour- substrate for the OT-mediated processing of motivation and aged by OT in a context-specific fashion. vigilance state for guiding social behavior. To date, most of the OT research examining the role of the amygdala in social cognition has focused on, broadly speaking, 8. Caveats in developing OT therapy for social social state modulations. Human functional neuroimaging disorders studies have shown that inhaling OT reduces BOLD (blood- oxygen-level dependent) activation in the amygdala when Without doubt, more research is necessary to fully and safely viewing or associating aversive social stimuli (Kirsch, 2005; develop and implement OT treatment protocols for social Petrovic et al., 2008), during trust formation (Baumgartner deficits. In particular, determining the efficacy and safety of et al., 2008), in response to pain (Singer et al., 2008), and in chronic OT treatment is a top priority because the majority of response to infant crying (Riem et al., 2011). Furthermore, OT OT studies to date have focused on establishing the effects of inhalation reduces amygdala activation when viewing fearful acute OT on specific social functions. Unfortunately, chronic faces but increases amygdala activation when viewing happy studies are practically difficult and sometimes ethically faces (Gamer et al., 2010). These findings suggest that OT questionable in human subjects. For these reasons, we modulates context-specific processing associated with both believe that non-human primates provide suitable models negative and positive valence in the amygdala, consistent with for monitoring the long-term safety and efficacy of repeated the encoding of both positive and negative motivational values OT administration, both behaviorally and physiologically. by the primate amygdala neurons (Belova et al., 2007; Paton Using non-human primates, researchers can administer OT et al., 2006). For mediating fear-related responses, it has been at a regular interval (e.g., daily or weekly) in the same verified in rats that oxytocinergic neurons regulate autonomic subjects over a long time period (e.g., months or years) to brain research 1580 (2014) 57–68 65

estimate the efficacy and safety of repeated OT treatments in be valuable for monitoring the efficacy and safety of repeated humans. It is also feasible to record neuronal activity or OT administrations continuously for months and even years measure hemodynamic responses regularly over time in such at the behavioral as well as neuronal level in a controlled studies in order to evaluate efficacy and safety at the setting. neurophysiological level. Furthermore, it is also feasible to conduct other physiological measurements every month to assess inflammatory processes, including urine and blood Author contributions analysis to measure cytokine and C-reactive protein levels. Another line of future research should focus on the S.W.C.C and M.L.P. wrote the paper. potential side effects of repeated exogenous OT exposure. In some cases, OT has been shown to evoke antisocial beha- viors, such as envy (Shamay-Tsoory et al., 2009), negative out- Conflict of interest group bias (De Dreu et al., 2011), negative social perception in individuals with anxious attachment (Bartz et al., 2010), The authors declare that they have no competing financial declined social interactions (Bales and Carter, 2003), as well interests. as aggressive behaviors, such as mate-guarding aggression displayed by female voles toward other females (Bales and Carter, 2003). These important studies support the idea that Acknowledgments the effects of OT depend critically on behavioral context (Bartz et al., 2011). Future studies thus need to determine This work was supported by K99-MH099093 (S.W.C.C.) and the neurobiological mechanisms responsible for antisocial or R01-MH086712 (M.L.P. and S.W.C.C.) from the National Insti- aggressive behaviors elicited by OT. Neuronal recording tute for Mental Health, and W81XWH-11-1-0584 from the studies in non-human primates designed to evoke their Department of Defense (M.L.P. and S.W.C.C.). natural prosocial or antisocial behaviors following OT infu- sion will provide important insights into the underlying references neural mechanisms, which can be refined by additional studies using molecular genetic techniques in rodents. It is also important to recognize the possibility that OT may Adolphs, R., 2010. What does the amygdala contribute to social acutely promote positive social behavior but may later pro- cognition?. Ann. N. Y. Acad. Sci. 1191, 42–61. Amaral, D.G., 2002. The primate amygdala and the neurobiology mote negative social behaviors after repeated or long-term of social behavior: implications for understanding social administration. anxiety. Biol. Psychiatry 51, 11–17. Amico, J.A., Mantella, R.C., Vollmer, R.R., Li, X., 2004. Anxiety and stress responses in female oxytocin deficient mice. J. 9. Concluding remarks Neuroendocrinol. 16, 319–324. Andari, E., Duhamel, J.-R., Zalla, T., Herbrecht, E., Leboyer, M., A rhesus macaque model of OT research has a clear transla- Sirigu, A., 2010. Promoting social behavior with oxytocin in high-functioning autism spectrum disorders. Proc. Natl. Acad. tional value for understanding and treating neuropsychiatric Sci. 107, 4389–4394. fi disorders with social de cits. Most importantly, it can provide Audigier, S., Barberis, C., 1985. Pharmacological characterization a unique opportunity for investigating how OT mediates of two specific binding sites for neurohypophyseal hormones human-like social cognition at both the neural circuit level in hippocampal synaptic plasma membranes of the rat. EMBO and the level of the whole animal. The social behaviors of J. 4, 1407–1412. rhesus macaques are remarkably similar to those of humans Averbeck, B.B., 2010. Oxytocin and the salience of social cues. Proc. Natl. Acad. Sci. 107, 9033–9034. (Maestripieri, 2007; Smuts, 1987). Arguably, most of the Bales, K.L., Carter, C.S., 2003. Sex differences and developmental rudimentary forms of complex human cognition are present effects of oxytocin on aggression and social behavior in prairie in these animals. Coupled with complex social behaviors, the voles (Microtus ochrogaster). Horm. Behav. 44, 178–184. amenability to single-unit recording and pharmacological Bandura, A., Ross, D., Ross, S.A., 1963. Vicarious reinforcement perturbations permits scientists to directly examine the and imitative learning. J. Abnorm. Psychol. 67, 601–607. neural mechanisms mediating OT influences on social cogni- Baron-Cohen, S., Wheelwright, S., Hill, J., Raste, Y., Plumb, I., 2001. tion. Furthermore, rhesus macaques are also tolerant to The “Reading the Mind in the Eyes” test revised version: a study with normal adults, and adults with asperger syndrome controlled experiments with an extremely high number of or high-functioning autism. J. Child Psychol. Psychiatry 42, repeated trials ( 1000 trials per day) allowing precise quan- 241–251. fi  ti cations of behaviors as well as how these behaviors are Barraza, J.A., McCullough, M.E., Ahmadi, S., Zak, P.J., 2011. related to underlying neuronal activity, which by nature is Oxytocin infusion increases charitable donations regardless of often highly variable. The efficacy of the nebulized inhalation monetary resources. Horm. Behav. 60, 148–151. method for delivering OT to the CNS in alert rhesus maca- Bartz, J.A., Zaki, J., Ochsner, K.N., Bolger, N., Kolevzon, A., Ludwig, ques prior to experimental tasks (Chang et al., 2012; Modi N., Lydon, J.E., 2010. Effects of oxytocin on recollections of maternal care and closeness. Proc. Natl. Acad. Sci. 107, et al., in press) makes it easier to compare neurophysiological 21371–21375. results from animal studies with behavioral or neuroimaging Bartz, J.A., Zaki, J., Bolger, N., Ochsner, K.N., 2011. Social effects of results obtained from both healthy people and those with oxytocin in humans: context and person matter. Trends Cogn. neuropsychiatric disorders. Finally, non-human primates will Sci. 15, 301–309. 66 brain research 1580 (2014) 57–68

Baumgartner, T., Heinrichs, M., Vonlanthen, A., Fischbacher, U., Cushing, B.S., Carter, C.S., 2000. Peripheral pulses of oxytocin Fehr, E., 2008. Oxytocin shapes the neural circuitry of trust increase partner preferences in female, but not male, prairie and trust adaptation in humans. Neuron 58, 639–650. voles. Horm. Behav. 37, 49–56. Baxter, M.G., Murray, E.A., 2002. The amygdala and reward. Nat. Deaner, R.O., Khera, A.V., Platt, M.L., 2005. Monkeys pay per view: Rev. Neurosci. 3, 563–573. adaptive valuation of social images by Rhesus Macaques. Curr. Belova, M.A., Paton, J.J., Morrison, S.E., Salzman, C.D., 2007. Biol. 15, 543–548. Expectation modulates neural responses to pleasant Dicks, D., Myers, R.E., Kling, A., 1969. Uncus and amiygdala and aversive stimuli in primate amygdala. Neuron 55, lesions: effects on social behavior in the free-ranging rhesus 970–984. monkey. Science 165, 69–71. Berber, S.M., 1962. Conditioning through vicarious instigation. Diester, I., Kaufman, M.T., Mogri, M., Pashaie, R., Goo, W., Yizhar, Psychol. Rev. 69, 450–466. O., Ramakrishnan, C., Deisseroth, K., Shenoy, K.V., 2011. An Bliss-Moreau, E., Bauman, M.D., Amaral, D.G., 2011. Neonatal optogenetic toolbox designed for primates. Nat. Neurosci. 14, amygdala lesions result in globally blunted affect in adult 387–397. Domes, G., Heinrichs, M., Michel, A., Berger, C., Herpertz, S.C., rhesus macaques. Behav. Neurosci. 125, 848–858. 2007. Oxytocin improves “Mind-Reading” in humans. Biol. Boccia, M.L., Petrusz, P., Suzuki, K., Marson, L., Pedersen, C.A., Psychiatry 61, 731–733. 2013. Immunohistochemical localization of oxytocin receptors Domes, G., Lischke, A., Berger, C., Grossmann, A., Hauenstein, K., in human brain. Neuroscience 253, 155–164. Heinrichs, M., Herpertz, S.C., 2010. Effects of intranasal Born, J., Lange, T., Kern, W., McGregor, G.P., Bickel, U., Fehm, H.L., oxytocin on emotional face processing in women. 2002. Sniffing neuropeptides: a transnasal approach to the Psychoneuroendocrinology 35, 83–93. human brain. Nat. Neurosci. 5, 514–516. Donaldson, Z.R., Young, L.J., 2008. Oxytocin, vasopressin, and the Brent, L.J.N., Chang, S.W.C., Garie´py, J.-F., Platt, M.L., neurogenetics of sociality. Science 322, 900–904. Neuroethology of friendship. Year Cogn. Neurosci., http://dx. Dreu, C.K.W.D., Greer, L.L., Handgraaf, M.J.J., Shalvi, S., Kleef, G.A. doi.org/10.1111/nyas.12315, in press. V., Baas, M., Velden, F.S.T., Dijk, E.V., Feith, S.W.W., 2010. Brent, L.J.N., Heilbronner, S.R., Horvath, J.E., Gonzalez-Martinez, J., The neuropeptide oxytocin regulates parochial altruism Ruiz-Lambides, A., Robinson, A.G., Skene, J.H.P., Platt, M.L., in intergroup conflict among humans. Science 328, 2013. Genetic origins of social networks in rhesus macaques. 1408–1411. Sci. Rep. 3. De Dreu, C.K.W., 2012. Oxytocin modulates cooperation within Byrne, R.W., Whiten, A., 1988. Machiavellian Intelligence: Social and competition between groups: an integrative review and Expertise and the Evolution of Intellect in Monkeys, Apes, and research agenda. Horm. Behav. 61, 419–428. Humans. Clarendon Press; Oxford University Press, Oxford; De Dreu, C.K.W., Greer, L.L., Van Kleef, G.A., Shalvi, S., Handgraaf, New York. M.J.J., 2011. Oxytocin promotes human ethnocentrism. Proc. Caffe´, A.R., Van Ryen, P.C., Van Der Woude, T.P., Van Leeuwen, Natl. Acad. Sci. USA 108, 1262–1266. F.W., 1989. Vasopressin and oxytocin systems in the brain and Dunbar, R.I., 1998. The social brain hypothesis. Brain 9, 10. upper spinal cord of Macaca fascicularis. J. Comp. Neurol. 287, Dunbar, R.I.M., Shultz, S., 2007. Evolution in the social brain. 302–325. Science 317, 1344–1347. Carter, C.S., 1998. Neuroendocrine perspectives on social Ebitz, R.B., Watson, K.K., Platt, M.L., 2013. Oxytocin blunts social attachment and love. Psychoneuroendocrinology 23, 779–818. vigilance in the rhesus macaque. Proc. Natl. Acad. Sci. 110, Champagne, F., Diorio, J., Sharma, S., Meaney, M.J., 2001. Naturally 11630–11635. occurring variations in maternal behavior in the rat are Feldman, R., Weller, A., Zagoory-Sharon, O., Levine, A., 2007. associated with differences in estrogen-inducible central Evidence for a neuroendocrinological foundation of human oxytocin receptors. Proc. Natl. Acad. Sci. 98, 12736–12741. affiliation: plasma oxytocin levels across pregnancy and the Champoux, M., Byrne, E., DeLizio, R., Suomi, S.J., 1992. Motherless postpartum period predict mother-infant bonding. Psychol. mothers revisited: Rhesus maternal behavior and rearing Sci. 18, 965–970. history. Primates 33, 251–255. Ferrari, P.F., Visalberghi, E., Paukner, A., Fogassi, L., Ruggiero, A., Chang, S.W.C., Winecoff, A.A., Platt, M.L., 2011. Vicarious Suomi, S.J., 2006. Neonatal imitation in Rhesus Macaques. reinforcement in rhesus macaques (Macaca Mulatta). Front. PLoS Biol. 4, e302. Flombaum, J.I., Santos, L.R., 2005. Rhesus monkeys attribute Neurosci. 5. Chang, S.W.C., Barter, J.W., Ebitz, R.B., Watson, K.K., Platt, M.L., perceptions to others. Curr. Biol. CB 15, 447–452. Francis, D.D., Champagne, F.C., Meaney, M.J., 2000. Variations in 2012. Inhaled oxytocin amplifies both vicarious reinforcement maternal behaviour are associated with differences in and self reinforcement in rhesus macaques (Macaca mulatta). oxytocin receptor levels in the rat. J. Neuroendocrinol. 12, Proc. Natl. Acad. Sci. 109, 959–964. 1145–1148. Chang, S.W.C., Garie´py, J.-F., Platt, M.L., 2013a. Neuronal reference Frith, C.D., Frith, U., 2007. Social cognition in humans. Curr. Biol. frames for social decisions in primate frontal cortex. Nat. 17, R724–R732. Neurosci. 16, 243–250. Galbally, M., Lewis, A.J., van IJzendoorn, M., Permezel, M., 2011. Chang, S.W.C., Brent, L.J.N., Adams, G.K., Klein, J.T., Pearson, J.M., The role of oxytocin in mother-infant relations: a systematic Watson, K.K., Platt, M.L., 2013b. Neuroethology of primate review of human studies. Harv. Rev. Psychiatry 19, 1–14. social behavior. Proc. Natl. Acad. Sci. USA 110 (Suppl. 2), Gamer, M., Zurowski, B., Buchel, C., 2010. Different amygdala 10387–10394. subregions mediate valence-related and attentional Choleris, E., Devidze, N., Kavaliers, M., Pfaff, D.W., 2008. effects of oxytocin in humans. Proc. Natl. Acad. Sci. 107, Steroidal/neuropeptide interactions in hypothalamus 9400–9405. and amygdala related to social anxiety. Prog. Brain Res. 170, Gimpl, G., Wiegand, V., Burger, K., Fahrenholz, F., 2002. 291–303. Cholesterol and steroid hormones: modulators of oxytocin Crockford, C., Wittig, R.M., Langergraber, K., Ziegler, T.E., receptor function. Prog. Brain Res. 139, 43–55. Zuberbu¨ hler, K., Deschner, T., 2013. Urinary oxytocin and Guastella, A.J., Mitchell, P.B., Dadds, M.R., 2008. Oxytocin social bonding in related and unrelated wild chimpanzees. increases gaze to the eye region of human faces. Biol. Proc. R. Soc. B Biol. Sci. 280. Psychiatry 63, 3–5. brain research 1580 (2014) 57–68 67

Guastella, A.J., Graustella, A.J., MacLeod, C., 2012. A critical review Miller, G., 2013. The promise and perils of oxytocin. Science 339, of the influence of oxytocin nasal spray on social cognition in 267–269. humans: evidence and future directions. Horm. Behav. 61, Modi, M.E., Connor-Stroud, F., Landgraf, R., Young, L.J., Parr, L.A., 410–418. 2013. Aerosolized oxytocin increases plasma and Harlow, H.F., Dodsworth, R.O., Harlow, M.K., 1965. Total social cerebrospinal fluid oxytocin in rhesus macaques. isolation in monkeys. Proc. Natl. Acad. Sci. USA 54, 90. Psychoneuroendocrinology (in press). Heinrichs, M., Domes, G., 2008. Neuropeptides and social Neumann, I.D., Torner, L., Wigger, A., 2000a. Brain oxytocin: behaviour: effects of oxytocin and vasopressin in humans. In: differential inhibition of neuroendocrine stress responses and Neumann, Inga D., Landgraf, Rainer (Eds.), Advances in anxiety-related behaviour in virgin, pregnant and lactating Vasopressin and Oxytocin – From Genes to Behaviour to rats. Neuroscience 95, 567–575. Disease, vol. 170. Elsevier, Oxford, UK, pp. 1–608. ISBN: 978-0- Neumann, I.D., Kro¨ mer, S.A., Toschi, N., Ebner, K., 2000b. Brain 444-53201-5. (Progress in Brain Research, 337–350). oxytocin inhibits the (re)activity of the hypothalamo- Heinrichs, M., Baumgartner, T., Kirschbaum, C., Ehlert, U., 2003. pituitary-adrenal axis in male rats: involvement of Social support and oxytocin interact to suppress cortisol and hypothalamic and limbic brain regions. Regul. Pept. 96, 31–38. subjective responses to psychosocial stress. Biol. Psychiatry Neumann, I.D., Maloumby, R., Beiderbeck, D.I., Lukas, M., 54, 1389–1398. Landgraf, R., 2013. Increased brain and plasma oxytocin after Hinde, R.A., 1974. Biological Bases of Human Social Behaviour. nasal and peripheral administration in rats and mice. McGraw-Hill, New York. Psychoneuroendocrinology 38, 1985–1993. Huber, D., 2005. Vasopressin and oxytocin excite distinct Nissenson, R., Flouret, G., Hechter, O., 1978. Opposing effects of neuronal populations in the central amygdala. Science 308, estradiol and progesterone on oxytocin receptors in rabbit 245–248. uterus. Proc. Natl. Acad. Sci. 75, 2044–2048. Van IJzendoorn, M.H., Huffmeijer, R., Alink, L.R.A., Bakermans- Parker, K.J., Buckmaster, C.L., Schatzberg, A.F., Lyons, D.M., 2005. Kranenburg, M.J., Tops, M., 2011. The impact of oxytocin Intranasal oxytocin administration attenuates the ACTH administration on charitable donating is moderated by stress response in monkeys. Psychoneuroendocrinology 30, experiences of parental love-withdrawal. Front. Psychol. 2. 924–929. Insel, T.R., 2010. The challenge of translation in social Parr, L.A., Modi, M., Siebert, E., Young, L.J., 2013. Intranasal neuroscience: a review of oxytocin, vasopressin, and affiliative oxytocin selectively attenuates rhesus monkeys’ attention to behavior. Neuron 65, 768–779. negative facial expressions. Psychoneuroendocrinology 38, Insel, T.R., Hulihan, T.J., 1995. A gender-specific mechanism for 1748–1756. pair bonding: oxytocin and partner preference formation in Paton, J.J., Belova, M.A., Morrison, S.E., Salzman, C.D., 2006. monogamous voles. Behav. Neurosci. 109, 782–789. The primate amygdala represents the positive and negative Insel, T.R., Shapiro, L.E., 1992. Oxytocin receptor distribution value of visual stimuli during learning. Nature 439, 865–870. reflects social organization in monogamous and polygamous Pedersen, C.A., 1997. Oxytocin control of maternal behavior. voles. Proc. Natl. Acad. Sci. 89, 5981–5985. Regulation by sex steroids and offspring stimuli. Ann. N. Y. Jenkins, J.S., Ang, V.T.Y., Hawthorn, J., Rossor, M.N., Iversen, L.L., Acad. Sci. 807, 126–145. 1984. Vasopressin, oxytocin and neurophysins in the human Pedersen, C.A., Ascher, J.A., Monroe, Y.L., Prange Jr, A.J., 1982. brain and spinal cord. Brain Res. 291, 111–117. Oxytocin induces maternal behavior in virgin female rats. Kemp, A.H., Guastella, A.J., 2010. Oxytocin: prosocial behavior, Science 216, 648–650. social salience, or approach-related behavior? Biol. Psychiatry Petrovic, P., Kalisch, R., Singer, T., Dolan, R.J., 2008. Oxytocin 67, e33. attenuates affective evaluations of conditioned Faces and Kirsch, P., 2005. Oxytocin modulates neural circuitry for amygdala activity. J. Neurosci. 28, 6607–6615. social cognition and fear in humans. J. Neurosci. 25, Rand, D.G., Greene, J.D., Nowak, M.A., 2012. Spontaneous giving 11489–11493. Kling, A., 1972. Effects of amygdalectomy on social-affective and calculated greed. Nature 489, 427–430. Riem, M.M.E., Bakermans-Kranenburg, M.J., Pieper, S., Tops, M., behavior in non-human primates, The Neurobiology of the Amygdala. Springer, US511–536. Boksem, M.A.S., Vermeiren, R.R.J.M., van IJzendoorn, M.H., Kluver, H., Bucy, P.C., 1939. Preliminary analysis of functions of Rombouts, S.A.R.B., 2011. Oxytocin modulates amygdala, the temporal lobes in monkeys. Arch. Neurol. Psychiatry 42, insula, and inferior frontal gyrus responses to infant crying: a 979 (–000). randomized controlled trial. Biol. Psychiatry 70, 291–297. Loup, F., Tribollet, E., Dubois-Dauphin, M., Dreifuss, J.J., 1991. Ring, R.H., Malberg, J.E., Potestio, L., Ping, J., Boikess, S., Luo, B., Localization of high-affinity binding sites for oxytocin and Schechter, L.E., Rizzo, S., Rahman, Z., Rosenzweig-Lipson, S., vasopressin in the human brain. An autoradiographic study. 2006. Anxiolytic-like activity of oxytocin in male mice: Brain Res. 555, 220–232. behavioral and autonomic evidence, therapeutic implications. MacDonald, K., Feifel, D., 2013. Helping oxytocin deliver: Psychopharmacology (Berl.) 185, 218–225. considerations in the development of oxytocin-based Rosvold, H.E., Mirsky, A.F., Pribram, K.H., 1954. Influence of therapeutics for brain disorders. Front. Neurosci. 7. amygdalectomy on social behavior in monkeys. J. Comp. Maestripieri, D., 2007. Macachiavellian intelligence: how rhesus Physiol. Psychol. 47, 173–178. macaques and humans have conquered the world. Sackett, G.P., 1984. A nonhuman primate model of risk for deviant Masserman, J.H., Wechkin, S., Terris, W., 1964. “Altruistic” development. Am. J. Ment. Defic. 88, 469–476. behavior in rhesus monkeys. Am. J. Psychiatry 121, 584–585. Santos, L.R., Nissen, A.G., Ferrugia, J.A., 2006. Rhesus monkeys, Meunier, M., Bachevalier, J., Murray, E.A., Ma´lkova´, L., Mishkin, M., Macaca mulatta, know what others can and cannot hear. Anim. 1999. Effects of aspiration versus neurotoxic lesions of the Behav. 71, 1175–1181. amygdala on emotional responses in monkeys. Eur. J. Sasaki, E., Suemizu, H., Shimada, A., Hanazawa, K., Oiwa, R., Neurosci. 11, 4403–4418. Kamioka, M., Tomioka, I., Sotomaru, Y., Hirakawa, R., Eto, T., et al., Meyer-Lindenberg, A., Domes, G., Kirsch, P., Heinrichs, M., 2011. 2009. Generation of transgenic non-human primates with Oxytocin and vasopressin in the human brain: social germline transmission. Nature 459, 523–527. neuropeptides for translational medicine. Nat. Rev. Neurosci. Saxe, R., 2006. Uniquely human social cognition. Curr. Opin. 12, 524–538. Neurobiol. 16, 235–239. 68 brain research 1580 (2014) 57–68

Shamay-Tsoory, S.G., Fischer, M., Dvash, J., Harari, H., Perach- Viviani, D., Charlet, A., van den Burg, E., Robinet, C., Hurni, N., Bloom, N., Levkovitz, Y., 2009. Intranasal administration of Abatis, M., Magara, F., Stoop, R, 2011. Oxytocin selectively oxytocin increases envy and schadenfreude (gloating). Biol. gates fear responses through distinct outputs from the central Psychiatry 66, 864–870. amygdala. Science 333, 104–107. Singer, T., Snozzi, R., Bird, G., Petrovic, P., Silani, G., Heinrichs, M., Waldherr, M., Neumann, I.D., 2007. Centrally released oxytocin Dolan, R.J., 2008. Effects of oxytocin and prosocial behavior on mediates mating-induced anxiolysis in male rats. Proc. Natl. brain responses to direct and vicariously experienced pain. Acad. Sci. 104, 16681–16684. Emotion 8, 781–791. Wang, Z., Moody, K., Newman, J.D., Insel, T.R., 1997. Vasopressin Smith, A.S., A˚ gmo, A., Birnie, A.K., French, J.A., 2010. and oxytocin immunoreactive neurons and fibers in the Manipulation of the oxytocin system alters social behavior forebrain of male and female common marmosets (Callithrix and attraction in pair-bonding primates, Callithrix penicillata. jacchus). Synapse 27, 14–25. Horm. Behav. 57, 255–262. Watson, K.K., Platt, M.L., 2012. Of mice and monkeys: using non- Smuts, B.B., 1987. In: Primate societies. University of Chicago human primate models to bridge mouse- and human-based Press, Chicago. investigations of autism spectrum disorders. J. Neurodev. Snowdon, C.T., Pieper, B.A., Boe, C.Y., Cronin, K.A., Kurian, A.V., Disord. 4, 21. Ziegler, T.E., 2010. Variation in oxytocin is related to variation Winslow, J.T., Insel, T.R., 1991. Social status in pairs of male in affiliative behavior in monogamous, pairbonded tamarins. squirrel monkeys determines the behavioral response to Horm. Behav. 58, 614–618. central oxytocin administration. J. Neurosci. 11, 2032–2038. Sofroniew, M.V., Weindl, A., Schrell, U., Wetzstein, R., 1981. Winslow, J.T., Hastings, N., Carter, C.S., Harbaugh, C.R., Insel, T.R., Immunohistochemistry of vasopressin, oxytocin and 1993. A role for central vasopressin in pair bonding in neurophysin in the hypothalamus and extrahypothalamic monogamous prairie voles. Nature 365, 545–548. regions of the human and primate brain. Acta Histochem. Winslow, J.T., Noble, P.L., Lyons, C.K., Sterk, S.M., Insel, T.R., 2003. Suppl. 24, 79–95. Rearing effects on cerebrospinal fluid oxytocin concentration Strathearn, L., Fonagy, P., Amico, J., Montague, P.R., 2009. Adult and social buffering in rhesus monkeys. attachment predicts maternal brain and oxytocin response to Neuropsychopharmacology 28, 910–918. infant cues. Neuropsychopharmacol. Off. Publ. Am. Coll. Wismer-Fries, A.B., Ziegler, T.E., Kurian, J.R., Jacoris, S., Pollak, Neuropsychopharmacol. 34, 2655–2666. S.D., 2005. Early experience in humans is associated with Subiaul, F., 2004. Cognitive Imitation in Rhesus Macaques. changes in neuropeptides critical for regulating social Science 305, 407–410. behavior. Proc. Natl. Acad. Sci. USA 102, 17237–17240. Suomi, S.J., 1991. Early stress and adult emotional reactivity in Yoshida, M., Takayanagi, Y., Inoue, K., Kimura, T., Young, L.J., rhesus monkeys. Child. Environ. Adult Dis., 171–188. Onaka, T., Nishimori, K., 2009. Evidence that oxytocin exerts Tribollet, E., Barberis, C., 1996. Vasopressin and Oxytocin anxiolytic effects via oxytocin receptor expressed in receptors in the central nervous system. Crit. Rev. Neurobiol. serotonergic neurons in mice. J. Neurosci. 29, 2259–2271. 10, 119–154. Young, L.J., 2002. The neurobiology of social recognition, Tribollet, E., Dubois-Dauphin, M., Dreifuss, J.J., Barberis, C., Jard, approach, and avoidance. Biol. Psychiatry 51, 18–26. S., 1992. Oxytocin Receptors in the Central Nervous System. Young, L.J., Wang, Z., 2004. The neurobiology of pair bonding. Nat. Ann. N. Y. Acad. Sci. 652, 29–38. Neurosci. 7, 1048–1054. Uvna¨s-Moberg, K., Ahlenius, S., Hillegaart, V., Alster, P., 1994. High Young, L.J., Huot, B., Nilsen, R., Wang, Z., Insel, T.R., 1996. Species doses of oxytocin cause sedation and low doses cause an differences in central oxytocin receptor gene expression: anxiolytic-like effect in male rats. Pharmacol. Biochem. Behav. comparative analysis of promoter sequences. J. 49, 101–106. Neuroendocrinol. 8, 777–783. brain research 1580 (2014) 160–171

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Oxytocin and the social brain: Neural mechanisms and perspectives in human research

Manuela Kanata,b, Markus Heinrichsa,b, Gregor Domesa,b,n aDepartment of Psychology, Laboratory for Biological and Personality Psychology, Albert-Ludwigs-University of Freiburg, D-79104 Freiburg, Germany bFreiburg Brain Imaging Center, University Medical Center, Albert-Ludwigs-University of Freiburg, D-79106 Freiburg, Germany article info abstract

Article history: The present paper summarizes functional imaging studies investigating the effects of Accepted 3 November 2013 intranasal oxytocin (OT) on brain responses to social stimuli. We aim to integrate previous Available online 8 November 2013 research, point to unresolved issues and highlight perspectives for future studies. The studies Keywords: so far have focused on identifying neural circuits underlying social information processing Functional imaging which are particularly sensitive to modulations by exogenous OT. Most consistently, Oxytocin stimulus-related responses of the amygdala and associated areas within the prefrontal and Brain activity temporal cortices have been found to be modulated by OT administration. However, there are Social behavior a number of unresolved issues related to the possible role of sex differences and hormonal Social cognition status, genetic variability, and individual differences in socio-cognitive functioning. Future studies focusing on these open questions are expected to contribute to a more nuanced understanding of the role of the central OT system in humans and may provide the basis for novel treatment approaches for mental disorders characterized by social deficits. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2013 Elsevier B.V. All rights reserved.

1. Introduction as pair bonding and parental care (e.g. Insel and Young, 2001; Young and Wang, 2004). 1.1. The role of oxytocin in human social cognition and In humans, most studies investigating the behavioral and behavior neural effects of OT have used placebo-controlled intranasal application. The rationale for this approach is based on Within the last decade, the neuropeptide oxytocin (OT) has findings that neuropeptides are capable of reaching the attracted increasing attention. Its importance for species- central nervous system following intranasal administration specific social functioning was first revealed by animal (Born et al., 2002). An actual increase in central OT levels studies demonstrating that central OT receptor distribution following intranasal OT administration was recently con- critically determines several aspects of social behavior such firmed by microdialysis in relevant brain regions of rats and

nCorresponding author at: Department of Psychology, Laboratory for Biological and Personality Psychology, Albert-Ludwigs-University of Freiburg, Stefan-Meier-Strasse 8, D-79104 Freiburg, Germany. Fax: 49 761 203 3023. þ E-mail address: [email protected] (G. Domes).

0006-8993/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.brainres.2013.11.003 brain research 1580 (2014) 160–171 161

mice (Neumann et al., in press). Studies that have explored the most commonly used brain imaging tool in human OT associations of peripheral OT levels with social stimulus research. processing are highly controversial regarding the validity of the assessment and interpretation with respect to CNS 1.3. Aims and structure of this article availability of the neuropeptide and need further investiga- tion (Anderson, 2006; Horvat-Gordon et al., 2005; Landgraf Here, we aim to provide a critical reflection on the current state and Neumann, 2004; Carter et al., 2007; for an overview, see of fMRI findings in OT research and highlight methodological Heinrichs et al., 2009). challenges and open questions that should be addressed by To date, effects of OT on human social cognition and behavior future studies. We will start withashortreviewofselectedfMRI have been summarized in several reviews and meta-analyses (e. studies integrating most of the previous evidence for effects of g. Heinrichs and Domes, 2008; Heinrichs et al., 2009; Shahrestani intranasal OT on neural correlates of social cognition and et al., 2013; Striepens et al., 2011). The main body of empirical behavior (Table 1). In the majority of these studies, a single evidence so far suggests beneficial effects of intranasal OT on dose of OT was administered to healthy individuals following a several aspects of social information processing and social randomized, placebo-controlled, experimental protocol. The behavior including eye gaze, facial emotion recognition, social cognitive functions under study included face perception and reward processing and trust (for recent reviews, see Guastella emotion processing (Domes et al., 2010a, 2007; Gamer et al., and MacLeod, 2012; Meyer-Lindenberg et al., 2011; Shahrestani 2010; Kirsch et al., 2005; Lischke et al., 2012; Petrovic et al., 2008; et al., 2013; Van IJzendoorn and Bakermans-Kranenburg, 2012). Rupp et al., 2012), proxies of parental sensitivity and attach- Together with results from animal studies, these findings point ment (Riem et al., 2012, 2011; Rupp et al., 2013; Wittfoth-Schardt to the therapeutic potential of interventions in the central et al., 2012), as well as different aspects of social feedback nervous OT system for the treatment of mental disorders processing (Baumgartner et al., 2008; Groppe et al., 2013; Rilling characterized by social impairments like social anxiety, schizo- et al., 2012). So far, few studies have focused on potential effects phrenia or autism spectrum disorders (for recent reviews, see of intranasal OT on brain responses in clinical samples char- Meyer-Lindenberg et al., 2011; Modi and Young, 2012; Striepens acterized by social deficits such as social anxiety (Labuschagne et al., 2011). Despite substantial evidence for beneficial effects of et al., 2011, 2010), borderline personality disorder (Bertsch et al., OT on social behavior, the neural mechanisms underlying these in press), or autism spectrum disorders (Domes et al., 2013a, in effects are still not well understood in humans. press; Watanabe et al., in press).

1.2. The oxytocin system in the human brain 2. Effects of OT on neural correlates of face perception and emotion processing In rats, OT receptors are highly expressed in olfactory and hypothalamic regions, structures of the (e.g. 2.1. The neural basis of social threat processing and amygdala), the thalamus, basal ganglia, as well as in the brain anxiolytic effects of OT stem and spinal cord (for a review, see Gimpl and Fahrenholz, 2001). OT binding sites in humans, however, remain rather Results from both animal studies and human behavioral elusive. Preliminary results suggest that OT receptor distribu- studies have consistently indicated anxiolytic effects of OT tion in the human brain differs substantially from other species (e.g. Bale et al., 2001; Heinrichs et al., 2003; Huber et al., 2005; with higher densities in dopaminergic neurons of the sub- Viviani et al., 2011). An initial fMRI study therefore focused on stantia nigra and the cholinergic nucleus basalis of Meynert OT effects on neural responses to threatening scenes and (Loup et al., 1991). However, Gimpl and Fahrenholz (2001) point facial expressions (Kirsch et al., 2005). Compared to placebo, to the possibility that high levels of OT in one region might intranasal administration of OT dampened amygdala reac- decrease local OT receptor expression to an extent that does tivity to negative social cues and reduced functional coupling not allow for detection by common methods like radioligand- of the amygdala with brainstem regions. These results based autoradiography. Thus, radioligands for in vivo PET correspond to recent findings from animal studies indicating studies in humans are still needed to provide a better under- that OT decreases behavioral fear responses by modulating standing of OT receptor distribution in the human brain. amygdala signaling to brainstem regions (Knobloch et al., While mapping of OT receptors in the human brain is still 2012; Viviani et al., 2011). in its infancy, functional brain imaging techniques allow for To gain a better understanding of potential anxiolytic OT mapping of brain regions potentially mediating the effects of effects in humans, a later fMRI study combined intranasal OT OT on human cognition and behavior. Some studies have administration with a fear-conditioning paradigm in which explored OT effects on brain activation using EEG or MEG neutral facial expressions were paired with electric shocks (Bick et al., 2013; Herzmann et al., in press; Hirosawa et al., (Petrovic et al., 2008). After conditioning, participants were 2012; Huffmeijer et al., 2013, 2012; Perry et al., 2010; Sheng administered a single dose of OT and underwent fMRI scanning et al., 2013). These methods have high temporal but rather during which they viewed the same faces with shock electrodes low spatial resolution, which complicates localization of applied. OT attenuated changes in affective stimulus ratings neural networks associated with the cognitive process under following fear conditioning andreducedneuralresponsesto study, especially if the brain areas of interest are located conditioned as compared to unconditioned stimuli in the subcortically. In comparison, spatial resolution is much amygdala, the medial temporal gyrus and the anterior cingulate higher in functional magnetic resonance imaging (fMRI), cortex. These structures are discussed as part of a neural alarm 162 brain research 1580 (2014) 160–171

Table 1 – Effects of intranasal oxytocin administration on brain responses to positive and negative stimulus exposure in healthy men and women.

Negative valence Positive valence

Sample Stimuli OT effects References Stimuli OT effects References

Male Faces and ↓ACC, AMY, CER, Kirsch et al. (2005), Domes Faces and ↓AMY Domes et al. (2007) scenes dlPFC, FG, ITL, et al. (2007), Petrovic et al. scenes ↑AMY Gamer et al. (2010) MED, MTL, OFC, (2008), Gamer et al. (2010), PAC, POC, PRC, Striepens et al. (2012), THAL, vlPFC Sauer et al. (2013) FC: AMY–BS, AMY–INS ↑INS Striepens et al. (2012) Social ↓ACC, AMY, BS, Baumgartner et al. (2008), Social ↑AMY, CAU Rilling et al. (2012) feedback CAU, INS, MB, Rilling et al. (2012) feedback POC ↑vlPFC, vmPFC Rilling et al. (2012) FC: INS-AMY, INS-IFG Physical pain ↓AMY, MB, STRIA Singer et al. (2008) ↑OFC Singer et al. (2008)

Female Faces and ↓AMY, dlPFC Domes et al. (2010a, Faces and ↓SMA Lischke et al. (2012) scenes 2010b), Rupp et al. (2012) scenes ↑AMY, ATL, BS, Domes et al. (2010a, ↑CER, FG, HIPP, Domes et al. CER, dlPFC, FG, 2010b), Lischke et al. INS, MTG, ROL, (2010a, 2010b) HIPP, INS, MTL, (2012) STG POC, ROL, STG, vlPFC Social ↓CUN, POC, PRC Groppe et al. (2013) Social ↑VTA, MTL, OCC Groppe et al. (2013) feedback feedback ↑VTA Groppe et al. (2013) Infant crying ↓AMY Riem et al. (2011) Infant laughter ↓AMY FC: AMY– Riem et al. (2012) OCC ↑IFG, INS Riem et al. (2011) ↑FC: AMY–ACC, Riem et al. (2012) AMY-ANG, AMY– HIPP, AMY–MTL, AMY–OFC, AMY– PREC

↑ increased brain activity, ↓ decreased brain activity, ACC anterior cingulate cortex, AMY amygdala, ANG angular gyrus, ATL anterior ¼ ¼ ¼ ¼ ¼ ¼ temporal lobe, BS brainstem, CAU caudate nucleus, CER cerebellum, CUN cuneus, dlPFC dorsolateral prefrontal cortex, FC functional ¼ ¼ ¼ ¼ ¼ ¼ connectivity, FG fusiform gyrus, HIPP hippocampus, IFL inferior frontal lobe, INS insula, ITL inferior temporal lobe, MB midbrain, ¼ ¼ ¼ ¼ ¼ ¼ MTL medial temporal lobe, MED medulla, OCC occipital cortex, OFC orbitofrontal cortex, PAC paracentral lobe, POC postcentral lobe, ¼ ¼ ¼ ¼ ¼ ¼ PRC precentral lobe, PREC precuneus, ROL rolandic operculum, SMA supplementary motor area, STG superior temporal gyrus, STRIA ¼ ¼ ¼ ¼ ¼ ¼ striatum, THAL thalamus, vlPFC ventrolateral prefrontal cortex, vmPFC ventromedial prefrontal cortex, and VTA ventral tegmental area. ¼ ¼ ¼ ¼

system in the context of threat processing (Liddell et al., 2005) task, OT reduced amygdala responses to emotional faces suggesting mitigation of the conditioned fear response follow- irrespective of valence (Domes et al., 2007), a result which is ing OT administration. Taken together, the studies by Kirsch at odds with the assumption that OT attenuates amygdala et al. (2005) and Petrovic et al. (2008) suggest that dampening responses specifically for threatening cues. Instead, the effects of OT on neural reactivity to threat-related social stimuli dampening effects of OT on amygdala reactivity to emotional might mediate its anxiolytic effects on human social behavior faces as found in our study may reflect a broader mechanism (Acheson et al., 2013; Heinrichs et al., 2003). such as decreased vigilance or uncertainty towards social events. 2.2. Differential effects of OT on emotional valence Such inconsistencies could also be explained by differen- processing and social attention tial interactions of emotion processing with visual attention processes. Accordingly, another study demonstrated that the OT effects within the amygdala, however, may not exclu- amygdala mediates OT effects on both social attention and sively be driven by negative stimuli. Within a passive-viewing emotional valence processing in a healthy male sample brain research 1580 (2014) 160–171 163

(Gamer et al., 2010). More specifically, OT increased the likelihood of reflexive saccades towards the eye regions of 3. Effects of OT on brain responses to infant faces independent of the displayed emotional expression. and sexual cues This main effect was associated with an increase in posterior amygdala activity. Discrimination of different emotional Several fMRI experiments examined effects of OT on respon- expressions, on the other hand, was found to be associated siveness to infant cues as a proxy of parental sensitivity and with anterior amygdala activity: OT attenuated activity for attachment. Healthy women were exposed to infant cry and fearful faces but increased activity for happy faces. A recent laughter sounds (Riem et al., 2012, 2011) and infant pictures study employing the same experimental paradigm in female (Rupp et al., 2013) following intranasal OT treatment. Under fi patients with borderline personality disorder showed that placebo, exposure to infant crying elicited signi cant activa- OT reduced patients' hypervigilance for threat cues and tion in the bilateral superior and middle temporal gyrus as associated abnormal amygdala reactivity (Bertsch et al., in well as the right amygdala (Riem et al., 2011). Intranasal OT press). Taken together, these results illustrate that OT effects reduced activation in the right amygdala but increased on visual attention should be taken into account when activation within the insula and the inferior frontal gyrus. studying its effects on social stimulus processing in the During exposure to infant laughter, OT once again attenuated amygdala. amygdala reactivity to emotional infant sounds, suggesting a mechanism that generalizes across positive and negative acoustic infant cues (Riem et al., 2012). OT also enhanced 2.3. Effects of oxytocin on emotion processing functional coupling of the amygdala with several brain in women regions associated with emotion processing and regulation, including the orbitofrontal cortex, anterior cingulate cortex, In an initial study, a sample of healthy women received hippocampus, and middle temporal gyrus. The results suggest intranasal OT or a placebo before viewing emotional or that OT might modulate fronto-cortical regulation of amygdala neutral face stimuli in an MRI scanner (Domes et al., 2010a). activation. In addition, decreased amygdala reactivity in post- Compared to placebo, OT increased reactivity to fearful partum women was also shown for the processing of infant facial expressions in brain areas involved in emotion proces- and sexual pictures (Rupp et al., 2013). A recent study on OT- sing, including the medial temporal lobe, amygdala, induced modulation of fathers' brain response to pictures of fusiform gyrus, superior temporal gyrus, and brainstem, but infants shows different results (Wittfoth-Schardt et al., 2012): reduced activation in the prefrontal cortex. Increased amyg- intranasal OT administration decreased left dala reactivity to negative stimuli after OT treatment in (GP) reactivity to pictures of the father's own child, and healthy women was confirmed in studies using threat- reduced functional coupling of the left GP with the right GP, related scenes (Lischke et al., 2012) and briefly presented the left hippocampus and the left middle frontal gyrus, facial expressions of anger (Bertsch et al., in press). The whereas amygdala activity appeared to be unaffected. increased amygdala reactivity under OT found in women contrasts with earlier studies in men which consistently reported attenuated amygdala responses following OT treat- 4. Modulation of neural activity by OT during ment. This leads to the hypothesis that OT may exert social feedback processing opposing effects on amygdala reactivity in healthy men and women. 4.1. Effects of OT on neural circuitry of trust and Moreover, effects of exogenous OT may differ depending reciprocity on individuals' endogenous OT levels, which are known to be elevated in postpartum women (Drewett et al., 1982; Uvnäs- In initial studies, OT was found to increase human trust Moberg et al., 1990). This hypothesis is supported behavior (Kosfeld et al., 2005), to maintain trust behavior by a recent fMRI study showing differential effects of OT following social betrayal, and to reduce neural responses on amygdala activation in response to negative pictures associated with the experience of breached trust in nulliparous and postpartum women (Rupp et al., 2012): (Baumgartner et al., 2008). Specifically, feedback of social nulliparous women had higher amygdala reactivity to betrayal increased activation within the amygdala and cau- negative stimuli than postpartum women under placebo; date nucleus under normal conditions but not following OT however, their amygdala reactivity was significantly administration (Baumgartner et al., 2008). Another study reduced by a single dose of intranasal OT, thereby making focused on reciprocal social cooperation in the context of a their neural responses comparable to those observed in prisoner's dilemma game (Rilling et al., 2012). Participants postpartum women. Thus, OT-induced attenuation of played with a putative human partner or a computer partner amygdala reactivity to arousing stimuli in women may who could either reciprocate or defect on their cooperation. reflect the neural basis for the stress-buffering effects of Compared to placebo, intranasal OT increased reactivity of breastfeeding in postpartum women (Mezzacappa and the left caudate nucleus and the left amygdala when partici- Katkin, 2002; for a review, see Heinrichs et al., 2002) and pants' cooperative behavior was mirrored by a putative known associations between breastfeeding, mother- human partner as compared to a computer partner. Activa- infant-bonding, and maternal care (Bosch and Neumann, tion in both regions was highly correlated, suggesting that OT 2012; Feldman et al., 2010, 2007; for a recent review, see increased the reward value of experienced cooperation or Galbally et al., 2011). enhanced the association of cooperative behavior with the 164 brain research 1580 (2014) 160–171

human partner. Together, these results support the hypoth- Asperger Syndrome (AS) and typically developed controls esis that OT shapes the processing of social cues in a (Domes et al., 2013a). Intranasal OT administration induced prosocial direction by increasing the reward value of positive preferential processing of faces within the amygdala in AS social cues while buffering against experiences of negative individuals which mirrored amygdala responses observed in emotionality. neurotypical controls under placebo. In another study, OT improved emotion recognition performance and heightened 4.2. Effects of OT on neural correlates of social reward and amygdala reactivity to emotional facial features (eyes and punishment processing mouths) in autistic individuals but not in typically developed controls (Domes et al., in press). A recent study directly tested whether OT influences the A recent study examined effects of intranasal OT on socio- neural circuitry of social saliency and reward in the context of communicational impairments in autism that stem from social feedback processing (Groppe et al., 2013). Female problems in interpreting nonverbal social cues (Watanabe participants performed a social incentive delay task within et al., in press). When simultaneously-presented verbal and the scanner in which they had to respond as quickly as nonverbal information was in conflict, OT increased the possible to a cued target stimulus in order to receive social number of judgments based on nonverbal as compared to reward (happy facial expressions) or avoid social punishment verbal information. For judgments based on nonverbal infor- (angry facial expressions). Under placebo, activation in the mation, OT enhanced activity in anterior cingulate and dorsal ventral tegmental area positively predicted correct responses medial prefrontal cortex – regions that were previously found in the social reward condition. OT increased activation in this to display reduced activity in autistic individuals as com- area during anticipation of both rewarding and punishing pared to typically developed controls (Watanabe et al., 2012). social cues. This result is in line with the hypothesis that OT Although a typically developed control sample was missing increases the motivational salience of social stimuli by in the study of Watanabe et al. (in press), the increase modulating neural correlates of reward and punishment in medial prefrontal cortex activity under OT may reflect processing. hormone-induced normalization of neural responses in autistic individuals. In sum, preliminary evidence suggests that OT may be capable of restoring neural correlates of social 5. Studies targeting clinical conditions and cognition in autism. individual differences 5.2. Normalizing effects of OT on neural hyperreactivity in Several mental disorders with impairments in social infor- social anxiety mation processing and social interaction are characterized by alterations in brain activation and connectivity. Such Socially anxious individuals typically show enhanced orient- alterations have recently been reviewed for schizophrenia ing towards socially threatening cues (Bar-Haim et al., 2007) (Fitzsimmons et al., 2013), autism (Philip et al., 2012), and and facilitated social threat conditioning (Pejic et al., 2013). social anxiety (Freitas-Ferrari et al., 2010). While beneficial These behavioral findings are mirrored in a neural hyper- effects of OT on social cognition and behavior have been reactivity during social threat processing with a prominent described for all of the above-mentioned neuropsychiatric role of the amygdala (Freitas-Ferrari et al., 2010; Pejic et al., disorders, recent meta-analytic results suggest that OT treat- 2013; Sladky et al., 2013). So far, two studies report modulat- ment may be specifically effective in autism spectrum dis- ing effects of OT on neural responses to emotional stimuli in orders (Bakermans-Kranenburg and van IJzendoorn, 2013a). social anxiety disorder (SAD) (Labuschagne et al., 2011, 2010): Overall, only few studies have explored the neural mechan- under placebo, individuals with clinical social anxiety as isms that might mediate effects of intranasal OT on social compared to healthy controls displayed a hyper-reactivity cognition and behavior in clinical samples. Clearly, pre- of the amygdala when processing facial expressions of fear existing differences in neural response patterns of clinical and an increased activity in the medial prefrontal cortex samples and healthy controls as well as individual drug (mPFC) and anterior cingulate cortex (ACC) when viewing sad medication need to be factored in when studying potential faces. Administration of OT reduced this neural hyper- therapeutic effects of OT. reactivity in social anxiety, supporting the assumption that OT may reduce stress reactivity and exert anxiolytic effects 5.1. OT-induced restoration of social cognition in autism by modulating activation in neural correlates of threat pro- cessing (Labuschagne et al., 2011, 2010). Autism spectrum disorders (ASD) are characterized by several socio-cognitive impairments, including difficulties in recog- 5.3. Influences of individual differences on reactivity to nizing faces or appropriately interpreting emotional cues of intranasal OT others. Previous evidence from behavioral studies suggests that OT exerts beneficial effects on social cognition and Evidence from several studies in healthy individuals suggests behavior in autism (Anagnostou et al., 2012; Andari et al., that OT effects may be generally moderated by individual 2010; Guastella et al., 2010). Based on these findings, recent levels of social functioning. For example, individuals with imaging studies focused on OT modulation of activity in the higher autistic traits indicative of poorer socio-cognitive social brain in autism. A first study tested for OT effects on functioning seem to show stronger beneficial effects of OT neural correlates of face processing in individuals with on empathic accuracy (Bartz et al., 2010). In female patients brain research 1580 (2014) 160–171 165

with borderline personality disorder, OT was found to However, OT research would generally benefit from a more decrease interpersonal trust and social cooperation preferen- thorough understanding of the neural networks that underlie tially in individuals with anxious attachment style and high fundamental cognitive processes that are assumed to be sensitivity towards social rejection (Bartz et al., 2011a). In modulated by exogenous OT. autistic individuals, an OT-induced increase in amygdala reactivity was predictive of individual improvements in behavioral performance (Domes et al., in press), suggesting 6.2. Accounting for connectivity within social brain that individual differences in social functioning may gener- networks ally account for differences in neural reactivity to intranasal OT within the social brain. Together, these findings illustrate To date, it remains unclear whether modulation of socio- the necessity of considering individual-related variables cognitive functions by OT is specifically or predominantly when studying effects of intranasal OT on behavioral and mediated by the amygdala. The studies so far simply suggest neural measures of socio-cognitive processing (compare that paradigms which are sensitive to modulations of amyg- Bartz et al., 2011b; Guastella and MacLeod, 2012; Kemp and dala activity seem to be particularly sensitive to changes in Guastella, 2011). central OT levels. Observed changes in amygdala responses following OT administration may therefore reflect local OT effects on other brain areas that exert modulatory influence 6. Methodological considerations for fMRI on amygdala activation. studies with pharmacological OT protocol This assumption addresses a crucial limitation of classic functional imaging. Functional MRI usually aims at identify- Imaging OT effects on brain activity associated with social ing brain areas involved in specific perceptual and cognitive stimulus processing has revealed several neural substrates processes. However, a purely localizing approach does not that may mediate its behavioral effects. Still, results from account for the existence of complex neural networks in fMRI studies remain heterogeneous, and caution is warranted which functions and performances arise from dynamic inter- in theorizing about oxytocinergic modulation of neural trans- actions between brain areas. Statistical tools for the analysis mission in humans. Functional MRI does not measure neural of connectivity measures allow for exploration of such activity per se, but instead measures alterations of magnetic regional interactions and usually focus either on modulations susceptibility in a brain area induced by temporary changes of functional vs. effective connectivity by an experimental of cerebral metabolism associated with neural activation. stimulation (Friston, 1994). This fact has some important implications. Two recent studies explored OT-induced alterations in functional connectivity between brain regions using resting 6.1. Choosing an appropriate experimental design state fMRI in men and women (Riem et al., in press; Sripada et al., 2013). In men, OT increased resting-state connectivity First, the size of experimental effects on the measured BOLD of the amygdala with anterior cingulate and medial prefron- response is generally low, and signal changes depend cru- tal cortex while reducing amygdala coupling with brainstem cially on the experimental paradigm. Inconsistent results on regions (Sripada et al., 2013). In women, resting state con- OT effects in the human brain as revealed by functional nectivity of the precuneus with the brainstem and the imaging may therefore arise from differences in task char- cerebellum was shifted in a positive direction following OT acteristics. These may include task demands (e.g. cognitive administration (Riem et al., in press). Together, preliminary load, implicit vs. explicit emotion processing), stimulus char- evidence from resting state fMRI points to modulatory effects acteristics (e.g. stimulus size, emotional intensity), timing of OT on brain networks involved in social cognition and issues (e.g. blocked vs. event-related design) and appropriate stress reduction. contrasting to an adequate control condition. In addition, Indeed, there is growing evidence for the assumption that neural responses to intranasally-administered OT may OT effects on social cognition and behavior are mediated by depend on the dosage used in a particular study. The transregional communication in the brain (for a recent review systematic exploration of dose–response characteristics of on OT effects on functional connectivity, see Bethlehem et al., intranasal OT administration should therefore be considered 2013). For example, increased attention to the eye region of an important next step in human OT research. faces following OT administration has been reported by both Especially if complex cognitive processes (such as trust) behavioral and neuroimaging studies (Domes et al., 2013b; and their modulation by OT are under study, interactions Gamer et al., 2010; Guastella et al., 2008). This increased eye with other cognitive processes are likely induced by an gaze seems to be mediated by enhanced functional coupling experimental task. As the amygdala was shown to display of the posterior amygdala with the superior colliculus under functional segregation, caution is warranted when interpret- OT (Gamer et al., 2010). OT was also found to increase ing OT effects in this brain region, as these can be mediated functional coupling of the amygdala with prefrontal and by different cognitive processes such as valence processing temporal regions as well as the hippocampus during proces- and attention orienting (see Gamer et al., 2010). In order sing of infant laughter (Riem et al., 2012) suggesting enhan- to address potential confounding effects of OT on social cing effects of OT on the neural processing of positive attention, eye movements can be assessed as a physiological stimulus valence (Iidaka et al., 2001; Whalen et al., 2013). correlate of visual attention processes during MRI scanning In contrast, reduced functional coupling of the amygdala with (Domes et al., 2013b; Gamer et al., 2010; Guastella et al., 2008). brainstem regions for aversive social stimulation may reflect 166 brain research 1580 (2014) 160–171

protective effects of OT against negative emotionality and (Derntl et al., 2008; Dreher et al., 2007; Marečková et al., stress reactivity (Kirsch et al., 2005; Rilling et al., 2012). 2012) and preliminary evidence suggests that brain morphol- It should be noted, however, that functional connectivity ogy itself may change across the menstrual cycle and follow- analyses describe statistical dependencies between the time- ing oral contraceptive use (Ossewaarde et al., 2013; Pletzer series of the measured BOLD response in different brain areas et al., 2010). and therefore represent mere correlations. To address the Evidence from animal studies indicates that the OT sys- question of causality within neural network communication tem is strongly influenced by gonadal hormones (Champagne and its modulation by OT, analysis of effective connectivity et al., 2001; De Kloet et al., 1986; Gabor et al., 2012; McCarthy, should be included in future research on OT effects on 1995; Tribollet et al., 1990). Although corresponding evidence human brain functioning (Friston et al., 2003; Stephan et al., for molecular cross-talk of steroid hormones with OT is 2010). lacking in humans, it seems likely that OT effects on human brain activity may interact with central actions of gonadal hormones. Studies investigating OT effects on brain activity 7. Sexual dimorphisms in neural correlates of in women therefore accounted for the menstrual cycle phase social cognition and implications for OT research (Bertsch et al., in press; Domes et al., 2010a; Lischke et al., 2012; Rupp et al., 2012). To date, a systematic exploration of 7.1. Sex differences in social processing and sexually potential effects of hormonal changes associated with the dimorphic effects of OT female menstrual cycle on reactivity to intranasal OT admin- istration is missing. Future studies addressing this issue may So far, results from fMRI studies exploring a modulation of contribute to a better understanding of the amount of social-stimulus related brain activity have suggested differ- variance in OT effects that might be explained by gonadal ential OT effects in men and women. These differences have hormones. In sum, there is pressing need of considering sex mainly been observed in the context of amygdala responses. hormone influences and menstrual cycle associated hormo- Studies in men have consistently reported attenuated amyg- nal changes in the study of central OT effects. dala responses following OT administration. In women, OT was found to increase amygdala reactivity to emotional cues in some studies (Domes et al., 2010a; Lischke et al., 2012) and to dampen amygdala activity in other studies (Riem et al., 8. OT and neurogenetics 2012, 2011; Rupp et al., 2012). The reasons for such differences may be manifold and no fMRI study has so far systematically 8.1. Association studies on the oxytocin receptor gene addressed the question of differential effects of OT on brain activity in men and women within a single experimental Given the high variance in behavioral and neural effects protocol. observed after OT application in most studies, one promising However, there is substantial evidence that men and women approach aims at identifying variations in specific genes differ in terms of brain anatomy and neural processing. For which contribute to individual differences in social behavior example, women have larger volumes of the orbitofrontal and social cognition, including vulnerability for neuropsy- cortex, the caudate and parts of the mirror-neuron system chiatric or developmental disorders characterized by social (Cheng et al., 2009; Filipek et al., 1994; Sowell et al., 2002); deficits (Ebstein et al., 2010). Several studies investigated whereas size of the hypothalamus, angular gyrus, and amyg- associations between the gene coding for the oxytocin recep- dala is larger in men than in women (Goldstein et al., 2001). tor (OXTR) and individual differences in social behavior (for a Notably, brain activity of men and women was found to differ review, see Kumsta and Heinrichs, 2013). on all domains of socio-cognitive functioning that show mod- The neurobiology underlying such associations between ulation by exogenous OT, e.g. emotion processing and regula- OXTR variants and social behavior phenotypes is addressed tion (Domes et al., 2010b;forrecentreviews,seeStevens and by the imaging genetics approach which relates genetic Hamann, 2012; Whittle et al., 2011). variants to brain structure and function (Meyer-Lindenberg et al., 2011). Specifically, imaging genetics aims at identifying 7.2. Effects of gonadal hormones on social stimulus endophenotypes (or intermediate phenotypes) such as altera- processing and potential interactions with OT tions in brain activity and morphometry that may bridge the gap between genotype and interpersonal variance in These sex differences are likely mediated by the early social behavior. Several neurogenetic studies showed that influence of gonadal hormones on differentiation and mor- genetic variation of OXTR affects a limbic circuit involving the phology of the central nervous system (Ahmed et al., 2008; amygdala, the hypothalamus and the cingulate gyrus (e.g. Schwarz and McCarthy, 2008; Sisk and Zehr, 2005) as well as Tost et al., 2011, 2010). Although preliminary data suggest their short-term modulation of neural transmission during that OXTR single nucleotide polymorphisms (SNPs) may rest and cognitive processing (Maki and Resnick, 2001). It is affect social cognition and behavior by modulating anatomy important to note that differences in gonadal hormone levels and functioning of the social brain (Kumsta and Heinrichs, in men and women (especially ) are not stable but 2013; Meyer-Lindenberg and Tost, 2012), meta-analytic evi- change with age as well as across the female menstrual cycle. dence for a direct impact of the two most frequently tested Menstrual cycle related changes were found to affect activity SNPs on human social behavior is missing (Bakermans- in brain networks underlying social stimulus processing Kranenburg and van Ijzendoorn, 2013b). brain research 1580 (2014) 160–171 167

8.2. Epigenetic and dopaminergic influences on central OT should further take variations of hormonal status over the signaling female menstrual cycle into account. In addition, systematic use of structural and functional genetic information may In addition to these structural changes, the importance of contribute to explain individual variance in brain responses epigenetic mechanisms that regulate genetic function and to exogenous OT. Differences in experimental design and the expression without affecting DNA structure was recently underlying cognitive processes addressed by a task add reviewed (Kumsta et al., 2013). In particular, methylation of further variability to fMRI findings on OT effects in the the OXTR promoter region is assumed to differentially influ- human brain. This issue relates to the fact that complex ence activity in brain regions associated with social percep- social stimulus processing involves a number of basic cogni- tion. Epigenetic states of genes are known to be modified by tive processes (perception, attention, memory, etc.). Future experiences, especially those occurring in sensitive periods imaging studies would benefit from a clear a priori definition early in life (e.g., traumatic experiences), forming the basis for of the basic cognitive processes under study and specifically a potential neurodevelopmental role of the OT system. More tailored experimental tasks. precisely, the influence of early adverse experiences on In addition, there is still a general lack of basic knowledge individual socio-emotional functioning may be mediated by regarding the distribution of OT receptors and the molecular epigenetically determined alterations of central oxytocin and cellular mechanisms of OT in the human brain. These signaling (Kumsta et al., 2013). issues are expected to be resolved as basic research into the Finally, two recent studies suggest that structural genetic mechanisms of neuropeptides and receptor mapping variations in genes regulating oxytocin and dopaminergic moves forward. As previous approaches to explore effects of signaling in the brain may interact with exogenous OT OT on brain connectivity were merely correlational, localizing administration (Sauer et al., 2013, 2012). It should be noted, the primary brain areas which are modulated by exogenous that the demand for large sample sizes to ensure reliability of OT may benefit from advanced methods for analyzing effec- observed interactions between genetic factors and exogenous tive connectivity using fMRI. Additionally, other neuroima- OT sets a severe restriction to such investigations. Still, future ging techniques, such as source-localization with high- fMRI studies may benefit from a neuropharmacogenetic density EEG may be employed. Together, these methods will approach as it surely helps to clarify the amount of variance broaden our search from a single brain area to functionally observed in measures of neuroanatomy and neural signaling integrated brain circuits underlying complex cognitive pro- that may be attributable to individual differences in genetic cesses, such as the recognition of an emotion from a facial factors. expression. In sum, advances made in functional and structural imaging techniques and neurogenetics over the past years 9. Conclusions and desiderata for future have revealed a yet-incomplete picture of the acute effects of research OT on brain function. However, future studies focusing on some of the issues highlighted in the present paper will In the past eight years, more than two dozen experimental undoubtedly help to reveal the neural functions of OT in the fMRI studies have been published regarding the effects of OT context of social information processing and might pave the on regional brain activity during social information proces- way to develop improved treatment strategies for mental sing, with many more currently underway. The studies so far disorders characterized by social dysfunction. show some variability in regard to the experimental task and the populations studied; nevertheless, most of the published studies have reported modulations of amygdala reactivity Acknowledgments in response to emotional as compared to neutral, or social compared to non-social, stimuli. The few fMRI studies We gratefully acknowledge the valuable discussion with in clinical samples suggest that this modulation depends Dr. Frances S. Chen. Preparation of the manuscript was on baseline socio-cognitive functioning of the population supported by a grant from the Deutsche Forschungsge- under study or differences in endogenous oxytocin signaling. meinschaft (DFG 1312/2-1). In addition, some studies have provided evidence for OT- induced modulations of functional coupling between the references amygdala and down-stream brain areas such as the brain- stem, as well as areas implicated in the regulation of amygdala reactivity such as parts of the prefrontal cortex. Acheson, D., Feifel, D., Wilde, S.de, Mckinney, R., Lohr, J., Presumably, differences between the studies summarized Risbrough, V., 2013. The effect of intranasal oxytocin above stem from two major sources of variance: (i) differ- treatment on conditioned fear extinction and recall in a ences between the populations under study (e.g. male vs. healthy human sample. Psychopharmacology. (Berl). 229, female and healthy vs. clinical samples) and (ii) differences in 199–208. regard to the primary cognitive processes involved in a task. Ahmed, E.I., Zehr, J.L., Schulz, K.M., Lorenz, B.H., DonCarlos, L.L., Sisk, C.L., 2008. Pubertal hormones modulate the addition of With regard to population characteristics, the issue of possi- new cells to sexually dimorphic brain regions. Nat. Neurosci. ble sex differences could be elucidated in future studies by 11, 995–997. directly comparing the effects of intranasal OT on male and Anagnostou, E., Soorya, L., Chaplin, W., Bartz, J., Halpern, D., female brain functions within the same study. These studies Wasserman, S., Wang, A.T., Pepa, L., Tanel, N., Kushki, A., 168 brain research 1580 (2014) 160–171

2012. Intranasal oxytocin versus placebo in the treatment of Cheng, Y., Chou, K.-H., Decety, J., Chen, I.-Y., Hung, D., Tzeng, O.J.-L., adults with autism spectrum disorders: a randomized Lin, C.-P., 2009. Sex differences in the neuroanatomy of human controlled trial. Mol. Autism 3, 1–9. mirror-neuron system: a voxel-based morphometric Andari, E., Duhamel, J.-R., Zalla, T., Herbrecht, E., Leboyer, M., investigation. Neuroscience 158, 713–720. Sirigu, A., 2010. Promoting social behavior with oxytocin in De Kloet, R., Voorhuis, D.A.M., Boschma, Y., Elands, J., 1986. high-functioning autism spectrum disorders. Proc. Natl. Acad. Estradiol modulates density of putative oxytocin receptors in Sci. 107, 4389–4394. discrete rat brain regions. Neuroendocrinology 44, 415–421. Anderson, G.M., 2006. Report of altered urinary oxytocin and AVP Derntl, B., Windischberger, C., Robinson, S., Lamplmayr, E., excretion in neglected orphans should be reconsidered. J. Kryspin-Exner, I., Gur, R.C., Moser, E., Habel, U., 2008. Facial Autism Dev. Disord. 36, 829–830. emotion recognition and amygdala activation are associated Bakermans-Kranenburg, M.J., van IJzendoorn, M.H., 2013a. A with menstrual cycle phase. Psychoneuroendocrinology 33, sociability gene? Meta-analysis of oxytocin receptor genotype 1031–1040. effects in humans. Psychiatr. Genet. http://dx.doi.org/10.1097/ Domes, G., Heinrichs, M., Gläscher, J., Büchel, C., Braus, D.F., YPG.0b013e3283643684. Herpertz, S.C., 2007. Oxytocin attenuates amygdala responses Bakermans-Kranenburg, M.J., van IJzendoorn, M.H., 2013b. to emotional faces regardless of valence. Biol. Psychiatry 62, Sniffing around oxytocin: review and meta-analyses of trials 1187–1190. in healthy and clinical groups with implications for Domes, G., Heinrichs, M., Kumbier, E., Grossmann, A., pharmacotherapy. Transl. Psychiatry 3, e258. Hauenstein, K., Herpertz, S.C., 2013a. Effects of intranasal Bale, T.L., Davis, A.M., Auger, A.P., Dorsa, D.M., McCarthy, M.M., oxytocin on the neural basis of face processing in autism 2001. CNS region-specific oxytocin receptor expression: spectrum disorder. Biol. Psychiatry 74, 164–171. importance in regulation of anxiety and sex behavior. Domes, G., Kumbier, E., Heinrichs, M., Herpertz, S.C., 2013. Oxytocin J. Neurosci. 21, 2546–2552. promotes facial emotion recognition and amygdala reactivity in Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, M.J., adults with asperger syndrome.Neuropsychopharmacology. van IJzendoorn, M.H., 2007. Threat-related attentional bias in http://dxdoi.org/10.1038/npp.2013.254. anxious and nonanxious individuals: a meta-analytic study. Domes, G., Lischke, A., Berger, C., Grossmann, A., Hauenstein, K., Psychol. Bull. 133, 1–24. Heinrichs, M., Herpertz, S.C., 2010a. Effects of intranasal Bartz, J.A., Simeon, D., Hamilton, H., Kim, S., Crystal, S., Braun, A., oxytocin on emotional face processing in women. Vicens, V., Hollander, E., 2011a. Oxytocin can hinder trust and Psychoneuroendocrinology 35, 83–93. cooperation in borderline personality disorder. Soc. Cogn. Domes, G., Schulze, L., Böttger, M., Grossmann, A., Hauenstein, K., Affect. Neurosci. 6, 556–563. Wirtz, P.H., Heinrichs, M., Herpertz, S.C., 2010b. The neural Bartz, J.A., Zaki, J., Bolger, N., Hollander, E., Ludwig, N.N., correlates of sex differences in emotional reactivity and Kolevzon, A., Ochsner, K.N., 2010. Oxytocin selectively emotion regulation. Hum. Brain Mapp. 31, 758–769. improves empathic accuracy. Psychol. Sci. 21, 1426–1428. Domes, G., Steiner, A., Porges, S.W., Heinrichs, M., 2013b. Bartz, J.A., Zaki, J., Bolger, N., Ochsner, K.N., 2011b. Social effects Oxytocin differentially modulates eye gaze to naturalistic of oxytocin in humans: context and person matter. Trends social signals of happiness and anger. Cogn. Sci. 15, 301–309. Psychoneuroendocrinology 38, 1198–1202. Baumgartner, T., Heinrichs, M., Vonlanthen, A., Fischbacher, U., Dreher, J.-C., Schmidt, P.J., Kohn, P., Furman, D., Rubinow, D., Fehr, E., 2008. Oxytocin shapes the neural circuitry of trust and Berman, K.F., 2007. Menstrual cycle phase modulates reward- trust adaptation in humans. Neuron 58, 639–650. related neural function in women. Proc. Natl. Acad. Sci. USA Bertsch, K., Gamer, M., Schmidt, B., Schmidinger, I., Walther, S., 104, 2465–2470. Kästel, T., Schnell, K., Büchel, C., Domes, G., Herpertz, S.C., Drewett, R.F., Bowen-Jones, A., Dogterom, J., 1982. Oxytocin levels 2013. Oxytocin and reduction of social threat hypersensitivity during breast-feeding in established lactation. Horm. Behav. in women with borderline personality disorder. Am. J. 16, 245–248. Psychiatry 170, 1169–1177. Ebstein, R.P., Israel, S., Chew, S.H., Zhong, S., Knafo, A., 2010. Bethlehem, R.A.I., van Honk, J., Auyeung, B., Baron-Cohen, S., Genetics of human social behavior. Neuron 65, 831–844. 2013. Oxytocin, brain physiology, and functional connectivity: Feldman, R., Gordon, I., Schneiderman, I., Weisman, O., a review of intranasal oxytocin fMRI studies. Zagoory-Sharon, O., 2010. Natural variations in maternal Psychoneuroendocrinology 38, 962–974. and paternal care are associated with systematic changes Bick, J., Dozier, M., Bernard, K., Grasso, D., Simons, R., 2013. Foster in oxytocin following parent–infant contact. mother–infant bonding: associations between foster mothers' Psychoneuroendocrinology 35, 1133–1141. oxytocin production, electrophysiological brain activity, Feldman, R., Weller, A., Zagoory-Sharon, O., Levine, A., 2007. feelings of commitment, and caregiving quality. Child Dev. 84, Evidence for a neuroendocrinological foundation of human 826–840. affiliation plasma oxytocin levels across pregnancy and the Born, J., Lange, T., Kern, W., McGregor, G.P., Bickel, U., Fehm, H.L., postpartum period predict mother–infant bonding. Psychol. 2002. Sniffing neuropeptides: a transnasal approach to the Sci. 18, 965–970. human brain. Nat. Neurosci. 5, 514–516. Filipek, P.A., Richelme, C., Kennedy, D.N., Caviness, V.S., 1994. Bosch, O.J., Neumann, I.D., 2012. Both oxytocin and vasopressin The young adult human brain: an MRI-based morphometric are mediators of maternal care and aggression in rodents: analysis. Cereb. Cortex 4, 344–360. from central release to sites of action. Horm. Behav. 61, Fitzsimmons, J., Kubicki, M., Shenton, M.E., 2013. Review of 293–303. functional and anatomical brain connectivity findings in Carter, S.C., Pournajafi-Nazarloo, H., Kramer, K.M., Ziegler, T.E., schizophrenia. Curr. Opin. Psychiatry 26, 172–187. White-Traut, R., Bello, D., Schwertz, D., 2007. Oxytocin: Freitas-Ferrari, M.C., Hallak, J.E.C., Trzesniak, C., Filho, A.S., behavioral associations and potential as a salivary biomarker. Machado-de-Sousa, J.P., Chagas, M.H.N., Nardi, A.E., Crippa, J.A.S., Ann. NY Acad. Sci. 1098, 312–322. 2010. Neuroimaging in social anxiety disorder: a systematic Champagne, F., Diorio, J., Sharma, S., Meaney, M.J., 2001. review of the literature. Prog. Neuropsychopharmacol. Biol. Naturally occurring variations in maternal behavior in the rat Psychiatry 34, 565–580. are associated with differences in estrogen-inducible central Friston, K.J., 1994. Functional and effective connectivity in oxytocin receptors. Proc. Natl. Acad. Sci. USA 98, 12736–12741. neuroimaging: a synthesis. Hum. Brain Mapp. 2, 56–78. brain research 1580 (2014) 160–171 169

Friston, K.J., Harrison, L., Penny, W., 2003. Dynamic causal Huffmeijer, R., Alink, L.R.A., Tops, M., Grewen, K.M., Light, K.C., modelling. NeuroImage 19, 1273–1302. Bakermans-Kranenburg, M.J., van IJzendoorn, M.H., 2013. The Gabor, C.S., Phan, A., Clipperton-Allen, A.E., Kavaliers, M., impact of oxytocin administration and maternal love Choleris, E., 2012. Interplay of oxytocin, vasopressin, and sex withdrawal on event-related potential (ERP) responses to hormones in the regulation of social recognition. Behav. emotional faces with performance feedback. Horm. Behav. 63, Neurosci. 126, 97–109. 399–410. Galbally, M., Lewis, A.J., van IJzendoorn, M., Permezel, M., 2011. Iidaka, T., Omori, M., Murata, T., Kosaka, H., Yonekura, Y., The role of oxytocin in mother-infant relations: a systematic Okada, T., Sadato, N., 2001. Neural interaction of the amygdala review of human studies. Harv. Rev. Psychiatry. 19, 1–14. with the prefrontal and temporal cortices in the processing of Gamer, M., Zurowski, B., Buchel, C., 2010. Different amygdala facial expressions as revealed by fMRI. J. Cogn. Neurosci. 13, subregions mediate valence-related and attentional effects of 1035–1047. oxytocin in humans. Proc. Natl. Acad. Sci. USA 107, 9400–9405. Insel, T.R., Young, L.J., 2001. The neurobiology of attachment. Nat. Gimpl, G., Fahrenholz, F., 2001. The oxytocin receptor system: Rev. Neurosci. 2, 129–136. structure, function, and regulation. Physiol. Rev. 81, 629–683. Kemp, A.H., Guastella, A.J., 2011. The role of oxytocin in Goldstein, J.M., Seidman, L.J., Horton, N.J., Makris, N., human affect: a novel hypothesis. Curr. Dir. Psychol. Sci. 20, Kennedy, D.N., Caviness, V.S., Faraone, S.V., Tsuang, M.T., 222–231. 2001. Normal sexual dimorphism of the adult human Kirsch, P., Esslinger, C., Chen, Q., Mier, D., Lis, S., Siddhanti, S., brain assessed by in vivo magnetic resonance imaging. Gruppe, H., Mattay, V.S., Gallhofer, B., Meyer-Lindenberg, A., Cereb. Cortex 11, 490–497. 2005. Oxytocin modulates neural circuitry for social cognition Groppe, S.E., Gossen, A., Rademacher, L., Hahn, A., Westphal, L., and fear in humans. J. Neurosci. 25, 11489–11493. Gründer, G., Spreckelmeyer, K.N., 2013. Oxytocin influences Knobloch, H.S., Charlet, A., Hoffmann, L.C., Eliava, M., Khrulev, S., processing of socially relevant cues in the ventral tegmental Cetin, A.H., Osten, P., Schwarz, M.K., Seeburg, P.H., Stoop, R., area of the human brain. Biol. Psychiatry 74, 172–179. Grinevich, V., 2012. Evoked axonal oxytocin release in the Guastella, A.J., Einfeld, S.L., Gray, K.M., Rinehart, N.J., Tonge, B.J., central amygdala attenuates fear response. Neuron 73, Lambert, T.J., Hickie, I.B., 2010. Intranasal oxytocin improves 553–566. emotion recognition for youth with autism spectrum Kosfeld, M., Heinrichs, M., Zak, P.J., Fischbacher, U., Fehr, E., 2005. disorders. Biol. Psychiatry 67, 692–694. Oxytocin increases trust in humans. Nature 435, 673–676. Guastella, A.J., MacLeod, C., 2012. A critical review of the influence Kumsta, R., Heinrichs, M., 2013. Oxytocin, stress and social of oxytocin nasal spray on social cognition in humans: behavior: neurogenetics of the human oxytocin system. Curr. evidence and future directions. Horm. Behav. 61, 410–418. Opin. Neurobiol. 23, 11–16. Guastella, A.J., Mitchell, P.B., Dadds, M.R., 2008. Oxytocin Kumsta, R., Hummel, E., Chen, F.S., Heinrichs, M., 2013. Epigenetic increases gaze to the eye region of human faces. Biol. regulation of the oxytocin receptor gene: implications for Psychiatry 63, 3–5. behavioral neuroscience. Front. Neurosci. http://dx.doi.org/10. Heinrichs, M., Baumgartner, T., Kirschbaum, C., Ehlert, U., 2003. 3389/fnins.2013.00083. Social support and oxytocin interact to suppress cortisol and Labuschagne, I., Phan, K.L., Wood, A., Angstadt, M., Chua, P., subjective responses to psychosocial stress. Biol. Psychiatry Heinrichs, M., Stout, J.C., Nathan, P.J., 2010. Oxytocin 54, 1389–1398. attenuates amygdala reactivity to fear in generalized social Heinrichs, M., Domes, G., 2008. Neuropeptides and social anxiety disorder. Neuropsychopharmacology. 35, 2403–2413. behaviour: effects of oxytocin and vasopressin in humans. Labuschagne, I., Phan, K.L., Wood, A., Angstadt, M., Chua, P., Prog. Brain Res. 170, 337–350. Heinrichs, M., Stout, J.C., Nathan, P.J., 2011. Medial frontal Heinrichs, M., Neumann, I., Ehlert, U., 2002. Lactation and stress: hyperactivity to sad faces in generalized social anxiety protective effects of breast-feeding in humans. Stress 5, disorder and modulation by oxytocin. Int. J. 195–203. Neuropsychopharmacol. 15, 883–896. Heinrichs, M., von Dawans, B., Domes, G., 2009. Oxytocin, Landgraf, R., Neumann, I.D., 2004. Vasopressin and oxytocin vasopressin, and human social behavior. Front. release within the brain: a dynamic concept of multiple and Neuroendocrinol. 30, 548–557. variable modes of neuropeptide communication. Front. Herzmann, G., Bird, C.W., Freeman, M., Curran, T., 2013. Effects of Neuroendocrinol. 25, 150–176. oxytocin on behavioral and ERP measures of recognition Liddell, B.J., Brown, K.J., Kemp, A.H., Barton, M.J., Das, P., memory for own-race and other-race faces in women and Peduto, A., Gordon, E., Williams, L.M., 2005. A direct men. Psychoneuroendocrinology 38, 2140–2151. brainstem–amygdala–cortical “alarm” system for subliminal Hirosawa, T., Kikuchi, M., Higashida, H., Okumura, E., Ueno, S., signals of fear. NeuroImage 24, 235–243. Shitamichi, K., Yoshimura, Y., Munesue, T., Tsubokawa, T., Lischke, A., Gamer, M., Berger, C., Grossmann, A., Hauenstein, K., Haruta, Y., Nakatani, H., Hashimoto, T., Minabe, Y., 2012. Heinrichs, M., Herpertz, S.C., Domes, G., 2012. Oxytocin Oxytocin attenuates feelings of hostility depending on increases amygdala reactivity to threatening scenes in emotional context and individuals' characteristics. Sci. Rep. females. Psychoneuroendocrinology 37, 1431–1438. http://dx.doi.org/10.1038/srep00384. Loup, F., Tribollet, E., Dubois-Dauphin, M., Dreifuss, J.J., 1991. Horvat-Gordon, M., Granger, D.A., Schwartz, E.B., Nelson, V.J., Localization of high-affinity binding sites for oxytocin and Kivlighan, K.T., 2005. Oxytocin is not a valid biomarker when vasopressin in the human brain. An autoradiographic study. measured in saliva by immunoassay. Physiol. Behav. 84, Brain. Res. 555, 220–232. 445–448. Maki, P.M., Resnick, S.M., 2001. Effects of estrogen on patterns of Huber, D., Veinante, P., Stoop, 2005. Vasopressin and oxytocin brain activity at rest and during cognitive activity: a review of excite distinct neuronal populations in the central amygdala. neuroimaging studies. NeuroImage 14, 789–801. Science 308, 245–248. Marečková, K., Perrin, J.S., Khan, I.N., Lawrence, C., Dickie, E., Huffmeijer, R., Alink, L.R.A., Tops, M., Bakermans-Kranenburg, M.J., McQuiggan, D.A., Paus, T., 2012. Hormonal contraceptives, IJzendoorn, M.H. van, 2012. Asymmetric frontal brain menstrual cycle and brain response to faces. Soc. Cogn. Affect. activity and parental rejection predict altruistic behavior: Neurosci. moderation of oxytocin effects. Cogn. Affect. Behav. McCarthy, M.M., 1995. Estrogen modulation of oxytocin and its Neurosci. 12, 382–392. relation to behavior. Adv. Exp. Med. Biol. 395, 235–245. 170 brain research 1580 (2014) 160–171

Meyer-Lindenberg, A., Domes, G., Kirsch, P., Heinrichs, M., 2011. and COMT gene variants influences the impact of oxytocin on Oxytocin and vasopressin in the human brain: social amygdala activation to social stimuli. Front. Neurosci. http:// neuropeptides for translational medicine. Nat. Rev. Neurosci. dx.doi.org/10.3389/fnins.2013.00045. 12, 524–538. Sauer, C., Montag, C., Wörner, C., Kirsch, P., Reuter, M., 2012. Meyer-Lindenberg, A., Tost, H., 2012. Neural mechanisms of Effects of a common variant in the CD38 gene on social social risk for psychiatric disorders. Nat. Neurosci. 15, 663–668. processing in an oxytocin challenge study: possible links to Mezzacappa, E.S., Katkin, E.S., 2002. Breast-feeding is associated autism. Neuropsychopharmacology 37, 1474–1482. with reduced perceived stress and negative mood in mothers. Schwarz, J.M., McCarthy, M.M., 2008. Steroid-induced sexual Health. Psychol. 21, 187–193. differentiation of the developing brain: multiple pathways, Modi, M.E., Young, L.J., 2012. The oxytocin system in drug one goal. J. Neurochem. 105, 1561–1572. discovery for autism: animal models and novel therapeutic Shahrestani, S., Kemp, A.H., Guastella, A.J., 2013. The impact of a strategies. Horm. Behav. 61, 340–350. single administration of intranasal oxytocin on the Neumann, I.D., Maloumby, R., Beiderbeck, D.I., Lukas, M., recognition of basic emotions in humans: a meta-analysis. Landgraf, R., 2013. Increased brain and plasma oxytocin after Neuropsychopharmacology 38, 1929–1936. nasal and peripheral administration in rats and mice. Sheng, F., Liu, Y., Zhou, B., Zhou, W., Han, S., 2013. Oxytocin Psychoneuroendocrinology 38, 1985–1993. modulates the racial bias in neural responses to others' Ossewaarde, L., van Wingen, G.A., Rijpkema, M., Bäckström, T., suffering. Biol. Psychol. 92, 380–386. Hermans, E.J., Fernández, G., 2013. Menstrual cycle-related Singer, T., Snozzi, R., Bird, G., Petrovic, P., Silani, G., Heinrichs, M., changes in amygdala morphology are associated with Dolan, R.J., 2008. Effects of oxytocin and prosocial behavior on changes in stress sensitivity. Hum. Brain. Mapp. 34, 1187–1193. brain responses to direct and vicariously experienced pain. Pejic, T., Hermann, A., Vaitl, D., Stark, R., 2013. Social anxiety Emotion 8, 781–791. modulates amygdala activation during social conditioning. Sisk, C.L., Zehr, J.L., 2005. Pubertal hormones organize the Soc. Cogn. Affect. Neurosci. 8, 267–276. adolescent brain and behavior. Front. Neuroendocrinol. 26, Perry, A., Bentin, S., Shalev, I., Israel, S., Uzefovsky, F., Bar-On, D., 163–174. Ebstein, R.P., 2010. Intranasal oxytocin modulates EEG mu/ Sladky, R., Höflich, A., Küblböck, M., Kraus, C., Baldinger, P., alpha and beta rhythms during perception of biological Moser, E., Lanzenberger, R., Windischberger, C., 2013. motion. Psychoneuroendocrinology 35, 1446–1453. Disrupted effective connectivity between the amygdala and Petrovic, P., Kalisch, R., Singer, T., Dolan, R.J., 2008. Oxytocin orbitofrontal cortex in social anxiety disorder during emotion attenuates affective evaluations of conditioned faces and discrimination revealed by dynamic causal modeling for fMRI. amygdala activity. J. Neurosci. 28, 6607–6615. Cereb. Cortex. http://dxdoi.org/10.1093/cercor/bht279. Philip, R.C.M., Dauvermann, M.R., Whalley, H.C., Baynham, K., Sowell, E.R., Trauner, D.A., Gamst, A., Jernigan, T.L., 2002. Lawrie, S.M., Stanfield, A.C., 2012. A systematic review and Development of cortical and subcortical brain structures in meta-analysis of the fMRI investigation of autism spectrum childhood and adolescence: a structural MRI study. Dev. Med. disorders. Neurosci. Biobehav. Rev. 36, 901–942. Child. Neurol. 44, 4–16. Pletzer, B., Kronbichler, M., Aichhorn, M., Bergmann, J., Sripada, C.S., Phan, K.L., Labuschagne, I., Welsh, R., Nathan, P.J., Ladurner, G., Kerschbaum, H.H., 2010. Menstrual cycle and Wood, A.G., 2013. Oxytocin enhances resting-state hormonal contraceptive use modulate human brain structure. connectivity between amygdala and medial frontal cortex. Int. Brain Res. 1348, 55–62. J. Neuropsychopharmacol. 16, 255–260. Riem, M.M.E., Bakermans-Kranenburg, M.J., Pieper, S., Tops, M., Stephan, K.E., Penny, W.D., Moran, R.J., den Ouden, H.E.M., Boksem, M.A.S., Vermeiren, R.R.J.M., van IJzendoorn, M.H., Daunizeau, J., Friston, K.J., 2010. Ten simple rules for dynamic Rombouts, S.A.R.B., 2011. Oxytocin modulates amygdala, causal modeling. NeuroImage 49, 3099–3109. insula, and inferior frontal gyrus responses to infant crying: a Stevens, J.S., Hamann, S., 2012. Sex differences in brain activation randomized controlled trial. Biol. Psychiatry 70, 291–297. to emotional stimuli: a meta-analysis of neuroimaging Riem, M.M.E., van IJzendoorn, M.H., Tops, M., Boksem, M.A.S., studies. Neuropsychologia 50, 1578–1593. Rombouts, S.A.R.B., Bakermans-Kranenburg, M.J., 2013. Striepens, N., Kendrick, K.M., Maier, W., Hurlemann, R., 2011. Oxytocin effects on complex brain networks are moderated by Prosocial effects of oxytocin and clinical evidence for its experiences of maternal love withdrawal. Eur. therapeutic potential. Front. Neuroendocrinol. 32, 426–450. Neuropsychopharmacol. 23, 1288–1295. Striepens, N., Scheele, D., Kendrick, K.M., Becker, B., Schäfer, L., Riem, M.M.E., van IJzendoorn, M.H., Tops, M., Boksem, M.A.S., Schwalba, K., Reul, J., Maier, W., Hurlemann, R., 2012. Oxytocin Rombouts, S.A.R.B., Bakermans-Kranenburg, M.J., 2012. No facilitates protective responses to aversive social stimuli in laughing matter: intranasal oxytocin administration changes males. Proc. Natl. Acad. Sci. USA 109, 18144–18149. functional brain connectivity during exposure to infant Tost, H., Kolachana, B., Hakimi, S., Lemaitre, H., Verchinski, B.A., laughter. Neuropsychopharmacology 37, 1257–1266. Mattay, V.S., Weinberger, D.R., Meyer–Lindenberg, A., 2010. Rilling, J.K., DeMarco, A.C., Hackett, P.D., Thompson, R., Ditzen, B., A common allele in the oxytocin receptor gene (OXTR) Patel, R., Pagnoni, G., 2012. Effects of intranasal oxytocin and impacts prosocial temperament and human hypothalamic- vasopressin on cooperative behavior and associated brain limbic structure and function. Proc. Natl. Acad. Sci. USA 107, activity in men. Psychoneuroendocrinology 37, 447–461. 13936–13941. Rupp, H.A., James, T.W., Ketterson, E.D., Sengelaub, D.R., Tost, H., Kolachana, B., Verchinski, B.A., Bilek, E., Goldman, A.L., Ditzen, B., Heiman, J.R., 2012. Amygdala response to negative Mattay, V.S., Weinberger, D.R., Meyer-Lindenberg, A., 2011. images in postpartum vs nulliparous women and intranasal Neurogenetic effects of OXTR rs2254298 in the extended oxytocin. Soc. Cogn. Affect. Neurosci. http://dx.doi.org/10. limbic system of healthy caucasian adults. Biol. Psychiatry 70, 1093/scan/nss100. e37–e39. Rupp, H.A., James, T.W., Ketterson, E.D., Sengelaub, D.R., Tribollet, E., Audigier, S., Dubois-Dauphin, M., Dreifuss, J.J., 1990. Ditzen, B., Heiman, J.R., 2013. Lower sexual interest in Gonadal steroids regulate oxytocin receptors but not postpartum women: relationship to amygdala activation and vasopressin receptors in the brain of male and female rats. An intranasal oxytocin. Horm. Behav. 63, 114–121. autoradiographical study. Brain Res. 511, 129–140. Sauer, C., Montag, C., Reuter, M., Kirsch, P., 2013. Imaging Uvnäs-Moberg, K., Widström, A.-M., Werner, S., Matthiesen, A.-S., oxytocin dopamine interactions: an epistasis effect of CD38 Winberg, J., 1990. Oxytocin and prolactin levels in breast- Â brain research 1580 (2014) 160–171 171

feeding women. Correlation with milk yield and duration of Kawakubo, Y., Matsuzaki, H., Tsuchiya, K.J., Kato, N., Kano, Y., breast-feeding. Acta Obstet. Gynecol. Scand. 69, 301–306. Miyashita, Y., Kasai, K., Yamasue, H., 2012. Diminished medial Van IJzendoorn, M.H., Bakermans-Kranenburg, M.J., 2012. A sniff prefrontal activity behind autistic social judgments of of trust: meta-analysis of the effects of intranasal oxytocin incongruent information. PLOS ONE 7, e39561, http://dx.doi.org/ administration on face recognition, trust to in-group, and 10.1371/journal.pone.0039561. trust to out-group. Psychoneuroendocrinology 37, 438–443. Whalen, P.J., Raila, H., Bennett, R., Mattek, A., Brown, A., Taylor, J., Viviani, D., Charlet, A., van den Burg, E., Robinet, C., Hurni, N., van Tieghem, M., Tanner, A., Miner, M., Palmer, A., 2013. Abatis, M., Magara, F., Stoop, R., 2011. Oxytocin selectively Neuroscience and facial expressions of emotion: the role of gates fear responses through distinct outputs from the central amygdala-prefrontal interactions. Emot. Rev. 5, 78–83. amygdala. Science 333, 104–107. Whittle, S., Yücel, M., Yap, M.B.H., Allen, N.B., 2011. Sex Watanabe, T., Abe, O., Kuwabara, H., Yahata, N., Takano, Y., differences in the neural correlates of emotion: evidence from Iwashiro, N., Natsubori, T., Aoki, Y., Takao, H., Kawakubo, Y., neuroimaging. Biol. Psychol. 87, 319–333. Kamio, Y., Kato, N., Miyashita, Y., Kasai, K., Yamasue, H., 2013. Wittfoth-Schardt, D., Gründing, J., Wittfoth, M., Lanfermann, H., Oxytocin mitigates socio-communicational deficits of autism Heinrichs, M., Domes, G., Buchheim, A., Gündel, H., Waller, C., with recovery of medial prefrontal activity: a randomized 2012. Oxytocin modulates neural reactivity to children's faces controlled trial. JAMA Psychiatry (in press). as a function of social salience. Neuropsychopharmacology Watanabe, T., Yahata, N., Abe, O., Kuwabara, H., Inoue, H., 37, 1799–1807. Takano, Y., Iwashiro, N., Natsubori, T., Aoki, Y., Takao, H., Young, L.J., Wang, Z., 2004. The neurobiology of pair bonding. Nat. Sasaki, H., Gonoi, W., Murakami, M., Katsura, M., Kunimatsu, A., Neurosci. 7, 1048–1054. brain research 1580 (2014) 8–21

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Oxytocin and vasopressin modulation of the neural correlates of motivation and emotion: results from functional MRI studies in awake rats

Marcelo Feboa,n, Craig F. Ferrisb,c,d aDepartment of Psychiatry, University of Florida McKnight Brain Institute, Gainesville, FL 32611, USA bCenter for Translational Neuroimaging, Northeastern University, Boston, MA 02115, USA cDepartment of Pharmaceutical Sciences, Northeastern University, Boston, MA 02115, USA dDepartment of Psychology, Northeastern University, Boston, MA 02115, USA article info abstract

Article history: Oxytocin and vasopressin modulate a range of species typical behavioral functions that Accepted 15 January 2014 include social recognition, maternal-infant attachment, and modulation of memory, Available online 30 January 2014 offensive aggression, defensive fear reactions, and reward seeking. We have employed Keywords: novel functional magnetic resonance mapping techniques in awake rats to explore the Functional MRI roles of these neuropeptides in the maternal and non-maternal brain. Results from the BOLD fMRI functional neuroimaging studies that are summarized here have directly and indirectly fi fi Rat con rmed and supported previous ndings. Oxytocin is released within the lactating rat fi Awake rat imaging brain during suckling stimulation and activates speci c subcortical networks in the Oxytocin maternal brain. Both vasopressin and oxytocin modulate brain regions involved uncondi- Vasopressin tioned fear, processing of social stimuli and the expression of agonistic behaviors. Across Maternal rat studies there are relatively consistent brain networks associated with internal motiva- Maternal attachment tional drives and emotional states that are modulated by oxytocin and vasopressin. Fear This article is part of a Special Issue entitled Oxytocin and Social Behav. & Anxiety 2014 Elsevier B.V. All rights reserved. Aggression Aggressive behavior Autism Addiction

1. Introduction recognition, modulation of social memory, maternal-infant attachment, offensive aggression, defensive and fear reac- It is well established that the neuropeptides oxytocin (OT) tions, and reward seeking behavior. Consistent with the and vasopressin (AVP) play major roles in what many range of neurobehavioral roles that have been discovered consider to be vital behavioral functions (Carter et al., for oxytocin and AVP, the receptor distribution for these 2008). The range of functions encompasses a spectrum of neuropeptides is observed to include various subcortical emotional and motivational mechanisms that include social forebrain and midbrain areas of rats, which subserve

nCorresponding author. E-mail addresses: [email protected], febo@ufl.edu (M. Febo).

0006-8993/$ - see front matter & 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.brainres.2014.01.019 brain research 1580 (2014) 8–21 9

the aforementioned behaviors. In non-human primates the Despite its advantages, it is important to keep in mind expression of their receptors may include regions of the shortfalls of fMRI in awake rats. First, the quantitative neocortex (Young et al., 1999). Among the challenges for assessment of excitatory/inhibitory neuronal firing as studying these integrative brain functions using a systems- recorded using other invasive techniques is limited in fMRI. level approach is the limited amount of non-invasive meth- Second, restraint used for fMRI experiments precludes beha- ods that are available. Our group has used functional mag- vioral tests during fMRI scanning. There are always concerns netic resonance mapping techniques to explore the about the effects of repeated exposure to transient, but modulatory roles of these neuropeptides in maternal and significant, levels of stress. This is an important and ongoing non-maternal brain networks. The present review offers a area of study by our laboratories. summary of these experiments in awake rats and puts forth several interpretations that could guide future fMRI studies in this area. 3. Affiliation: imaging oxytocin and maternal-offspring interactions

2. Functional magnetic resonance imaging OT acts within the lactating brain to provide a major signal- of awake rats ing mechanism that strengthens the mother-offspring bond early in life. Its synthesis mainly occurs in neurons of Functional magnetic resonance methods developed by Ferris and paraventricular (PVN) and supraoptic nucleus (SON) of colleagues have allowed fMRI studies in awake restrained rats hypothalamus. Suckling stimulates the release of OT into (Ferris et al., 2005). There are obvious advantages to this experi- the bloodstream via the neurohypophyseal portal system and mental paradigm that is often overlooked by many in the in the central nervous system of postpartum rats (Neumann neuroscience community. Most of the available fMRI studies in et al., 1993b). Systemically, OT stimulates smooth muscle rodents are carried out while animals are under anesthesia. This contraction, which is important for milk ‘let-down’ during is done despite the anticipation of studying neural circuits that nursing and for uterine contraction during parturition. OT underlie or drive distinct motivational and cognitive states. This, release within the CNS during parturition initiates maternal of course, is impossible when the awake state is suppressed by behaviors and may act to coordinate emotion, social and . Experiments that employ the awake rat methods cognitive networks necessary for maternal care. Many of the require acclimation to restraint and MR sound prior to experi- brain areas involved in maternal behavior in rat and other ments. We have been able to successfully implement these species of mammals express moderate-to-high levels of OT procedures in our studies. This requires the animals be placed receptors (Tribollet et al., 1988a; Tribollet et al., 1988b; Vaccari under transient stress over the span of 5 days or so. Results from et al., 1998; Veinante and Freund-Mercier, 1997; Yoshimura previous studies show promising outcomes in which animals et al., 1993), suggesting a convergence of maternal and OT show reductions in physiological and motor perturbations that pathways. Brain regions include amygdala, dorsal hippocam- may hinder the quality of the collected data (King et al., 2005). pus, hypothalamic paraventricular, ventromedial, and pre- Moreover, the behavioral outcomes of restraint acclimation for optic nuclei, lateral septum, olfactory structures, nucleus fMRI do not appear to be permanent (Reed et al., 2013a, 2013b). accumbens, substantia nigra (SN), ventral tegmental area Thus, the studies cited in the present review are all collected (VTA), bed nucleus of stria terminalis (BNST), among others while rats are in the awake, unanesthetized state. Details of the (See Fig. 1 and Table 1). Elevated OT receptor density in areas experimental paradigms used, especially pre-processing strate- such as the preoptic area (mPOA) has been correlated with gies to screen for and to minimize the impact of motion artifact greater levels of pup licking and grooming in postpartum have been published (Ferris et al., 2008). dams (Champagne et al., 2001; Francis et al., 2000). It is in The non-invasive measurement of neural activity through these structures that OT may contribute to strengthening of the blood oxygenation level dependent (BOLD) signal is key to mother-infant interactions. OT release in response to suck- the enthusiasm in pursuing fMRI studies rodents and other ling has been measured using microdialysis in the substantia species. To date, there are no neuroscience techniques that nigra, olfactory bulbs, mediobasal hypothalamus, BNST, can supersede and replace this feature of fMRI. The BOLD MPOA and septum of parturient ewes (Levy et al., 1995), as signal arises from changes in tissue oxy-to-deoxyhemoglobin well as in sites of origin, PVN and SON, of rat (Neumann et al., ratio and. Microvascular magnetic field gradients that are 1993a; Neumann et al., 1993b). Administration of an OT sensitive to changes in paramagnetic deoxy- and diamag- receptor antagonist locally within the latter two nuclei netic oxy-hemoglobin concentrations near brain areas of reduced OT release, suggesting that a positive feedback altered neuronal metabolism are a likely source of the BOLD mechanism is involved (Neumann et al., 1994). Moreover, fMRI signal. Changes in neuronal activity, and the accompa- administration of an OT antagonist locally within the mPOA nying compensatory adjustments in blood flow, in blood reduces arched back nursing in the rat (Bosch and Neumann, volume and in the cerebral consumption rates for oxygen, 2012). underlie the mapping of the measured BOLD signal changes. In one of the first experiments investigating the putative A variety of experimental paradigms have been developed functional neural circuits of lactating rat brain, we presented over the years in order to examine brain function with fMRI. fMRI evidence that suckling-induced OT release into central The premise in all of the paradigms is the external delivery of neural sites increases the BOLD signal. This effect on the a sensory-driving stimulus, be it drug, autonomic, primary BOLD signal was reported for PVN, olfactory tubercle, anterior sensory, or a higher order complex stimulus. olfactory nucleus (AON), insular cortex, piriform cortex, 10 brain research 1580 (2014) 8–21

cortical amygdala, mPOA and prefrontal cortex. For several of of brain areas co-localized with OT receptors, or that are asso- the regions, suckling and central OT administration increased ciated with OT-mediated behaviors (Table 1). Centrally adminis- the BOLD signal in a strikingly similar pattern. The promising tered OT activated areas associated with maternal, and other results were not unexpected if the technique were to reliably reproductive behaviors, e.g. the primary olfactory system, the show lactation neural circuits, and the contribution of OT release ventral medial hypothalamus, mPOA and BNST. The 3D color into these sites. In addition, central administration of an OT model on the top of Fig. 1 depicts the brain areas in the rat that 2 4 9 receptor antagonist (d(CH2)5-[Tyr(Me) ,Thr ,Tyr-NH2]-Ornithine are reported to display a high density of OT receptor binding (De Vasotocin) partly blocked the suckling-induced increase in the Kloet et al., 1985; Freund-Mercier et al., 1987; Tribollet et al., 1988c; BOLD signal in cortical, midbrain and olfactory regions (Febo van Leeuwen et al., 1985). As there are no reported differences in et al., 2005). OT antagonist blockade in olfactory regions suggests OT receptor binding sites between males and female rats (Barberis that OT release during nursing contributes to olfactory-related and Tribollet, 1996), this 3D model is representative of both sexes. neural activity. The results of the above-cited study (Febo et al., These areas have been coalesced into a single volume (yellow), as 2005)illustrateacommonthemepresentthroughseveralofthe shown in the lower 3D images for ICV administration. Areas in subsequent neuroimaging studies in rat. There is a close func- red are from composite images of BOLD signal changes collected tional interaction between OT and AVP with olfactory function. in nine rats, each given OT ICV (1.0 ug). The localization of BOLD This may be particularly important for enhancing the perception activation is clearly shown on the 2D axial sections of the MR atlas or facilitation of the registration of socially relevant stimuli into of rat. Blockade of OT receptors with a specificantagonist short- and long-term memory circuits of rats. selectively reduced brain activity in many of these areas, evidence We have observed that neuranatomical substrates activated by that endogenous OT has a role in the neurobiology of nursing. central OT administration in the lactating rat closely paralleled There were non-overlapping brain regions, namely the caudate- that observed with suckling stimulation alone. OT given intracer- putamen, septum, and thalamus that might involve other neuro- ebroventricularly (ICV) showed a dose-dependent BOLD activation transmitter systems aside from OT.

Fig. 1 – Neural circuitry of high density ot receptor binding sites. The 3D color model on the top depicts the brain areas in the rat with a high density of OT receptor. These areas are coalesced into a single volume (yellow) within a translucent shell of the brain, as shown in the lower 3D images for ICV. Areas colored red are the localization of activated voxels representing the composite average of nine rat brain fMRI scans. Once fully registered, and segmented, the statistical responses for each animal are averaged on a voxel-by-voxel basis. The averaged voxels that are significantly different from baseline, and exceed a 2.0% threshold, are show in their appropriate spatial location. The 2D images appearing to the right of each 3D image show the location of interpolated voxels of positive BOLD activation registered onto coronal sections of the segmented rat atlas. The approximate location of these brain slices are depicted by the black lines shown in the 3D ICV image. brain research 1580 (2014) 8–21 11

Table 1 – Brain activation with intraventricular oxytocin.

As noted above, there are numerous studies showing that induces maternal behavior in virgin rats (Yu et al., 1996). the main olfactory system, indeed the itself, is The memory preservation of social recognition of juvenile critical for processing and remembering odors that relevant female conspecifics is reduced with injection of OT receptor for social recognition (Sanchez-Andrade and Kendrick, 2009). antagonist into the olfactory bulb (Larrazolo-Lopez et al., Volatile chemical signals from the environment interact with 2008), while OT infused in the olfactory bulbs of male rats odorant receptors on the olfactory epithelium. In the rat, preserves the memory of prior social interactions (Dluzen there are approximately 1200 genes coding for odorant et al., 1998). Indeed, the olfactory bulb, and its downstream receptors (Quignon et al., 2005). These olfactory sensory connections to cortical amygdala, medial amygdala, BNST neurons project to mitral cells in the glomerular layer of the and medial preoptic areas, are all integrated in social recog- olfactory bulb. Olfactory information is processed at the level nition and all are sensitive to OT neurotransmission. of olfactory bulbs and conveyed to the olfactory cortex and amygdala. Natural maternal behavior in rats and mice (Fleming and Rosenblatt, 1974; Gandelman et al., 1971), maternal recognition of offspring in ewe (Baldwin and 4. Aggression: imaging vasopressin Shillito, 1974) and social recognition in male rats (Dantzer and intruder aggression et al., 1990), require the olfactory bulbs. OT plays a significant role in each. During parturition and vaginal-cervical stimula- 4.1. AVP V1a modulation of maternal neural correlates tion, OT levels increase in the olfactory bulbs (Kendrick et al., of aggression 1988a; Kendrick et al., 1988b; Larrazolo-Lopez et al., 2008). Blocking OT receptors in the olfactory bulbs immediately AVP varies from its cousin nonapeptide OT by way of after parturition delays maternal behavior in rats (Yu et al., substitution of the amino acids Phe for Ile in the 3rd position, 1996). Conversely, OT injections into the olfactory bulbs and Arg for Leu in the 8th position (Gimpl and Fahrenholz, 2001). 12 brain research 1580 (2014) 8–21

It is implicated in the regulation of numerous social and shown that central blockade of V1a receptors modulated behavioral processes such as aggression, social bonding, and BOLD signal responses in primiparous dams (Caffrey et al., maternal behavior (Goodson and Bass, 2001). Within the mam- 2010). The differential BOLD responses were not generalized fi malian central nervous system, the synaptic actions of AVP across the maternal rat brain, but were site speci c. V1a are mediated to a large degree by the V1a receptor subtype receptor antagonist enhanced the volume of BOLD activation (Tribollet et al., 1988a), which is found throughout the rodent in the AON, and reduced it in the cortical amygdala and VMH. brain (Ostrowski et al., 1994), and the V1b receptor subtype. Greater percentage increases in BOLD signal were observed

AVP is important for the modulation of aggressive behaviors across several brain areas in response to V1a antagonist in males and females. Ferris and Potegal, (1988) demonstrated treatment, including the AON, gustatory cortex, infralimbic that microinjection of a V1a receptor antagonist in the anterior area, substantia innominata (which expresses V1a receptors), hypothalamus reduced resident male hamster aggression and the somatosensory cortex. An unexpected finding of the fi toward a male intruder. Similarly, a microinjection of AVP into study was that V1a receptor blockade signi cantly enhanced the lateral ventricle causes an increased BOLD signal in regions BOLD signal in somatosensory areas during intruder presen- of the brain involved in aggression and that are known to tation. The greater BOLD signal response occurred in both contain V1a receptors (Ferris et al., 2008). In microtine rodents, primary and supplemental areas, but was not generalized to AVP also has been observed to promote offensive aggression, the entire cortical mantle. The heightened BOLD response as well as partner preference (Winslow et al., 1993). Both the observed in this region of the cortex was not observed in fi aggressive and af liative responses may be mediated by V1a primary or secondary motor cortical areas, or in parietal or receptors in this species of rodent (Lim et al., 2004; Young, 1999). temporal cortices. Thus, somatosensory modulation by V1a AVP is also important in maternal behavior, as chronic AVP receptors appears to be selective for this cortical region. The treatment in lactating rats increases maternal care (Bosch and only region that showed a lower percentage change in BOLD

Neumann, 2008), and V1a antagonists impair maternal memory with V1a receptor blockade was the VMH. The results indi-

(Nephew and Bridges, 2008)andreducenursingandpup cated that V1a receptors modulate neural processing in retrieval (Pedersen et al., 1994). specific neural circuits recruited during a timeframe corre- One specific type of maternal behavior is maternal aggres- sponding to the initial phases of maternal aggressive motiva- sion, which is a robust form of aggression most evident tion. The V1a sensitive circuits include structures which are within the first two weeks of lactation (Erskine et al., 1978). involved in sensory processing, the control of visceral This aggressive response can be eliminated by bilateral responses, and emotional memory. One interesting feature olfactory bulbectomy (Kolunie and Stern, 1995); however, of the brain areas that were modulated by V1a receptors in deprivation of auditory and visual inputs has no effect on our imaging studies is that some of these share connectivity maternal aggression (Kolunie et al., 1994), indicating that with the orbital prefrontal cortex in rats (Ongur and Price, olfaction, but not auditory or visual stimulation, is essential 2000). Gustatory, infralimbic, olfactory, somatosensory and for maternal aggression. Several brain regions have been amygdalar networks are processed through this limbic cor- associated with maternal aggression, including regions of tical structure that plays an important role in sensory and the limbic system such as the amygdala, nucleus accumbens visceromotor associations (Gabbott et al., 2005; Ongur and and BNST (Nephew et al., 2009; Numan and Numan, 1996). Price, 2000; Vertes, 2004). Although V1a receptors in rat are not The hypothalamus is also associated with the onset of localized in the various cortical regions, the observed enhance- maternal behavior, particularly the VMH (Bridges and Mann, ment might be achieved through indirect subcortical actions of 1994). Lesions to the VMH advance the onset of maternal this receptor antagonist. It remains to be determined whether or behavior in primigravid rats (Mann and Babb, 2004), suggest- not this is the case. Activation of the somatosensory cortex and ing that this region may be inhibitory toward maternal its modulation by V1a receptors is interesting in light of the fact behavior. Interestingly, anxiogenic state may influence how that AVP receptors have not been detected in this brain region, OT and AVP modulate maternal aggression. Low anxiety and that a somatosensory stimulus was not presented to dams. behavior (LAB) dams reportedly show lower basal levels of Thalamic sensory relay nuclei were unaffected by antagonist attacks towards a nest intruder than high anxiety behavior treatment, and do not contain V1a receptors. However, the (HAB) dams (Bosch and Neumann, 2012). Stimulating OT substantia innominata, which has been shown to contain an receptors in LAB dams increases the number of attacks, understudied population of V1a receptors (Ostrowski et al., 1994), whereas blocking OT receptors decreases maternal aggres- sends major cholinergic projections to the cortical mantle, in sion in HAB dams (Bosch and Neumann, 2012). Similarly, particular the somatosensory and prefrontal cortices. Lesions of blocking vasopressin V1a receptors also increases maternal these cholinergic projections differ dramatically from sensory aggression in LAB dams and reduces it in HAB dams (Bosch cortical ablations, and alter emotional reactivity in rats (Knox and Neumann, 2010). The similar outcomes of these two et al., 2008; Wozniak et al., 1989). Indeed, it is possible that V1a peptides might be the result of ‘cross-talk’ via their receptors. receptors in this basal forebrain area, and its interaction with Cortical regions involved in maternal aggression are less the somatosensory cortex, may control aspects of behavioral well understood. Conscious lactating female rats were reactivity, and perception of internal autonomic/visceral states. injected ICV with either a V1a antagonist or saline, and Due to the transient nature of the BOLD changes in the presented with a novel male intruder in the presence of her substantia innominata, it is postulated that this nucleus may pups. Brain activation was measured using BOLD fMRI to mediate the immediate response to the male intruder, perhaps determine the role of V1a receptors on maternal brain activa- mapping maternal autonomic bodily reactions to possible tion during presentation of a novel male intruder. It was aggression towards pups (Caffrey et al., 2010). Though the above brain research 1580 (2014) 8–21 13

interpretations appear plausible in light of previous literature regulated by 5-HT. The hypothalamus, the primary site of and our reported data, they remain highly speculative at the AVPergic facilitation of aggression, has a high density of 5- moment. HT1a and 5-HT1b binding sites and receives a dense innerva- tion of 5-HT fibers and terminals (Delville et al., 2000; Ferris

4.2. Male aggression: role of AVP V1a neurotransmission et al., 1997; Grimes and Melloni, 2002; Ricci et al., 2006). Hypothalamic AVP neurons implicated in the control of With regards to the actions of AVP on aggression in rat and in aggression appear to be preferentially innervated by 5-HT other rodent models, it appears to partly mediate its effect via (Ferris et al., 1991). Fluoxetine blocks resident-intruder interactions with serotonin (5-HT) at the synaptic level. There aggression that is facilitated by the microinjection of AVP in is a body of literature reporting that blockade of V1a receptors the hypothalamus (Delville et al., 1996; Ferris, 1996; Ferris in a variety of animal models suppresses aggression (Ferris, et al., 1997). Fluoxetine elevates 5-HT and reduces AVP levels

2005). Consequently, drugs that target, and block, the V1a in hypothalamic tissue in hamsters (Ferris, 1996) and rats receptors are being developed as potential therapeutics for (Altemus et al., 1992). Serotonin can also block the activity of the treatment of impulsivity and violence. Recently, a new AVP following its release in the hypothalamus as evidenced class of non-peptidic compounds targeting the human V1a by the dose-dependent reduction of aggression with injec- receptor was developed using a monocyclic beta lactam tions combining AVP and 5-HT1a receptor . Enhanced platform (Guillon et al., 2006). One of these potential drugs, aggression caused by activation of V1a receptors in the SRX251, was tested for serenic activity in the hamster (Ferris hypothalamus is suppressed by the simultaneous activation et al., 2006). It was shown that oral administration of SRX251 of 5-HT1a receptors in the same site (Ferris et al., 1997). Person- caused a dose-dependent decrease in several measures of ality disordered subjects with a history of fighting and assault aggressive behavior, without affecting motor activity, olfac- show a negative correlation for prolactin release in response to D- tory communication, and sexual motivation. fenfluramine challenge, indication of a hyposensitive 5-HT Precisely how and where 5-HT and AVP interact to affect system (Coccaro et al., 1998). These same subjects show a the organization, and expression, of aggressive behavior is positive correlation between CSF levels of vasopressin and unclear. Normal aggressive behaviors, and aggression char- aggression (Coccaro et al., 1998). Thus, in humans a hyposensi- acterized by impulsivity and violence, are envisioned to be tive 5-HT system may result in enhanced CNS levels of AVP, with organized and controlled by a distributed neural circuit. aconsequentfacilitationofaggressivebehavior. These include subsets of interconnected neurons conveying While fluoxetine and SRX251 have similar effects on the sensory and motor information to and from sites of integra- putative neural circuitry of aggressive motivation, a markedly tion (Ferris et al., 2008). With functional magnetic resonance different fMRI signature was observed with each compound imaging (fMRI) it has been possible to identify the neural when treated males were challenged with sexual motivating circuitry involved in aggressive motivation using an experi- stimuli (Ferris et al., 2008). With V1a receptor blockade there mental paradigm that drives an agonistic state just prior to was activation of the SN, VTA, and their afferent projections the onset of attack. As part of the ethogram of aggression, to the forebrain limbic cortex, as well as the dorsal and resident male rats housed with a female cage will piloerect ventral striatum. Measures of sexual activity in the home along the dorsal midline when in the presence of a male environment were unaffected by SRX251 treatment. Treat- intruder. The piloerection is unique to offensive aggression, ment with fluoxetine, on the other hand, resulted in a and is not seen in other behaviors signaling an impending diminished BOLD activation in response to sexual motivating attack (Blanchard and Blanchard, 1977). Using a novel aggres- stimuli. It also caused inhibition of sexual behavior in the sion model adapted to the neuroimaging conditions, we home environment. These opposite effects point to a differ- discovered that even though a resident male is confined to a ence in drug specificity, and underscore the serenic proper- restraining device for an imaging session, placing an intruder ties of SRX251, specifically its ability to block aggression into the vivarium with its cage mate induces piloerection. Our without affecting other appetitive behaviors. results identified distributed putative neural circuits asso- ciated with the genesis of attack behavior and their differential modulation but AVP and 5-HT acting compounds. 5. Anxiety: imaging oxytocin and vasopressin Enhanced 5HT neurotransmission is associated with a in unconditioned fear reduction in aggressive responding via interaction with 5-

HT1a and 5-HT1b receptors (Simon et al., 1998; Grimes and 5.1. OT and unconditioned fear in the lactating rat brain Melloni, 2005). Oral fluoxetine suppressed aggression and diminished BOLD activation across the putative neural circuit Attending and responding effectively to environmental dan- of aggressive motivation (Ferris et al., 2008). Conversely, AVP gers is important during the postpartum period. Lactating neurotransmission promotes aggression by interacting with rats show aggression to conspecific males and are generally

V1a receptors. Oral SRX251, a V1a receptors antagonist, sup- less fearful than virgin rats. Central release of OT and AVP pressed aggression and produced a general reduction in BOLD during lactation may contribute to heightened maternal activation in the neural circuitry of aggression similar to that aggression and nest defense, and lower fearfulness (Bale seen with fluoxetine (Ferris et al., 2008). The observation that et al., 2001; Blume et al., 2008; Bosch and Neumann, 2008; fluoxetine and SRX251 are similar in their fMRI profile during Francis et al., 2000; Windle et al., 1997a; Windle et al., 1997b). suppression of aggressive motivation was anticipated. There In the case of OT, it is released in the PVN and supraoptic is evidence that the stimulation of aggression by AVP is nucleus during lactation (Neumann et al., 1993b) and nest 14 brain research 1580 (2014) 8–21

defense (Bosch et al., 2005), respectively. Blocking OT recep- Our results for OT and fear are in agreement with a tors reduces lactation-related behaviors (Bosch and previous study looking at the BOLD response to OT adminis- Neumann, 2008; Pedersen and Boccia, 2003), increases anxi- tration in lactating rats (Febo et al., 2005). ICV OT administra- ety related behavior in an elevated plus maze (Bosch and tion was observed to activate the anterior cingulate, parietal Neumann, 2008), and affects maternal offensive and defen- cortex, somatosensory, temporal cortices, dorsal striatum, sive behaviors (Bosch et al., 2005). The actions of OT on preoptic area and ventral tegmental area. The actions of OT aggression and fear are perhaps exerted through PVN and administration on BOLD responses within these regions may central amygdala (Bosch et al., 2005; Heinrichs et al., 2008); hold importance to emotional reactivity, particularly to innate however, it is quite possible that a wider circuitry including fear reactions. For instance, the anterior cingulate is involved other limbic and forebrain sites contributes to OT's role in in the initiation of goal directed behaviors (Devinsky et al., fear reactivity during the postpartum period. 1995), as well as the emotional control of visceral, skeletal, and We investigated the effect of centrally administered OT on endocrine functions (Vogt et al., 1992). BOLD activation in response to an unconditioned fear stimulus OT receptor binding sites, and its receptor-encoding in dams. Trimethylthiazoline (TMT), a chemical extract of fox mRNA, have been detected in the BNST, lateral septum, feces, has been shown to evoke fear-associated freezing behavior posterior cortical amygdala, anterior olfactory nucleus, piri- (Wallace and Rosen, 2001), and it also has been observed to form cortex, main olfactory bulbs, central amygdala, subicu- elevate plasma levels of corticosterone (Chen et al., 2009; Febo lum, and dispersed cortical regions (Veinante and Freund- et al., 2009; Morrow et al., 2000). TMT-associated increase in Mercier, 1997; Yoshimura et al., 1993). The observed TMT- corticosterone in lactating rats is enhanced when pups are in the induced BOLD responses were in partial agreement with nest (Deschamps et al., 2003). The physiological and behavioral areas of the rat brain expressing OT receptors. The BNST is signs of stress and fear in response to TMT are distinguishable particularly interesting. This region is part of the extended from responses to novel acrid odors (Morrow et al., 2000; Staples amygdala and not only participates in modulating stress and et al., 2008), and considered innate since laboratory rodents have fear responses, but also expresses both OT and V1a receptors. no prior exposure to predator odors. We observed a greater volume of activation in this area with Our fMRI findings showed that OT treatment modulates OT pretreatment, which suggests a greater neuronal activity BOLD signal increases and decreases in response to TMT in in response to TMT. Recent report shows that TMT exposure lactating rats. OT enhanced TMT-induced positive BOLD elevates extracellular norepinephrine levels in the BNST signal changes in the anterior cingulate, a brain region (Fendt et al., 2005). Blocking norepinephrine receptors with involved in emotional expression and fear conditioning. clonidine results in less time spent freezing in response to While the BNST, which was previously shown to be respon- TMT (Fendt et al., 2005). Although electrolytic lesions of the sive to TMT (Day et al., 2004), and also constitute part of the lateral amygdala block conditioned fear, but not innate fear neural circuitry mediating behavioral responses to this pre- (Wallace and Rosen, 2001), the opposite has been shown for dator scent (Fendt et al., 2003; Fendt et al., 2005), also showed the BNST (Fendt et al., 2003). This suggests an essential role of greater BOLD signal as compared to vehicle controls. The the BNST in unconditioned fear response (Fendt et al., 2003). latter basal forebrain region is interesting in light of the fact It might be interesting to examine the interactions between that it also modulates VTA dopamine neuron activity noradrenergic inputs and OT inputs to this region. It should (Georges and Aston-Jones, 2001; Georges and Aston-Jones, be noted that the BNST also is important in behavioral and 2002), thus providing a neurobiological interface between endocrine responses to contextual fear stimuli (Sullivan et al., distinct motivation and emotional states. Interestingly, a 2004). Therefore, the BNST may be an important brain area wider array of structures showed negative BOLD signal where OT exerts its effect on TMT-induced freezing in lactating changes in response to TMT following OT administration. rats. Although there is no evidence showing OT receptors in the The anterior olfactory nucleus, which serves as an initial anterior cingulate, this region shares synaptic communication relay site for olfactory-processing, showed primarily signal with the septum and amygdala, which might indirectly have decreases. This was observed in the prelimbic prefrontal regulated TMT-induced BOLD activation. Indeed, previous cortex, orbital cortex, mammillary bodies, secondary motor reports indicate that the anterior cingulate cortex plays a role cortex and the gustatory cortex (Febo et al., 2009). The pattern in memory of emotional and fear-inducing events (Malin and of reduced brain activity (negative BOLD) underscores the McGaugh, 2006; Malin et al., 2007). importance of OT in regulating fear and anxiogenic responses during the lactation period. Moreover, the pattern varies from 5.2. OT and AVP modulation of unconditioned fear findings in male rats (see below). It is tempting to speculate in the male rat brain on whether this represents an underlying physiological mechanism where activity in the anterior cingulate during Brain regions that are associated with unconditioned fear exposure to an innate fear stimulus occurs in parallel with a partly overlap with receptors for the neuropeptides OT and lack of neuronal activity or inhibition in other adjacent limbic AVP. Therefore, it is important to consider the role of these cortical regions controlling approach behavior and autonomic neuromodulators as potential pharmaceutical targets for responses. Negative BOLD has, however, less support from treatment of generalized anxiety disorders. The central dis- the literature on its role in neuronal processing than positive tribution of OT and AVP suggest that they play a role in BOLD (Logothetis, 2003). Although somewhat controversial, modulating olfactory processing, social behavior, as well as there are indications that negative BOLD correlates with cognitive, emotion-related, and goal-directed behavioral reductions in synaptic activity (Shmuel et al., 2006). functions (Huber et al., 2005; Ostrowski et al., 1994; Szot brain research 1580 (2014) 8–21 15

et al., 1994; Tribollet et al., 1988a; Tribollet et al., 1988b; Using c-fos immunolabeling, Staples et al. (2008) identified Veinante and Freund-Mercier, 1997). There is growing evi- regions of rat brain that responded selectively to cat and TMT dence that OT-mediated effects result in anxiolysis, while over no odor and formalin odor. They reported selective cat-

AVP actions through the V1a receptor promote aggression and induced activation of c-fos labeling in VTA, striatum (dorsal heightened anxiety (Appenrodt et al., 1998; Knobloch et al., and ventral), basal and medial amygdaloid nuclei, subareas of 2012; Murgatroyd et al., 2004; Waldherr and Neumann, 2007). AON, medial prefrontal cortex, anterior and ventromedial Not only do these neuropeptides control anxiety, but also their hypothalamic nuclei, BNST and dorsal premammilary presence in amygdala and hippocampus suggests a prominent nucleus (Staples et al., 2008). TMT had only modest effects role in cognitive mechanisms, such as conditioned fear learning on cellular activation, increasing activity in the cortical processes (De Wied et al., 1975; de Wied et al., 1984; Vawter et al., amygdala (anterior portion), piriform and ventral orbital 1997). Based on their location within the components of the cortices. Rosen et al. (2005) reported increased mRNA expres- olfactory system, such as AON and the olfactory tubercles, as sion of egr-1 in the hypothalamic paraventricular nucleus well as central amygdala and BNST, these peptides may control (PVN) of rats following exposure to a cat (Rosen et al., 2005). olfactory perception and the association between smell, social This is interesting since this region contains OT and V1a recognition and emotionality (Ferris, 2008). receptors that control the central release of their correspond- We used functional MRI to investigate the role of OT and ing neuropeptides via magnocellular neurons (Knobloch

V1a receptors in modulating odor-evoked brain activity in the et al., 2012). Several lesion studies have provided evidence awake male rat (Reed et al., 2013a, 2013b). A novel repulsive of a role for the ventral hippocampus and BNST in unlearned odor for the rats, butyric acid (BA), and unknown chemical fear reactivity. Inactivation of the ventral BNST or blockade of constituents present in cat fur were chosen in order to norepinephrine α2 receptors reduces freezing in response to examine unconditioned fear and anxiety responses. The TMT (Fendt et al., 2003; Fendt et al., 2005). Similar effects are results indicated that BA-induced BOLD signal increases were observed with lateral septum inactivation (Endres and Fendt, fi signi cantly greater than with cat fur. This presented chal- 2008). These areas express V1a receptor mRNA (Veinante and lenges to the final interpretations of the data since the fear- Freund-Mercier, 1997). In the case of the ventral hippocam- eliciting chemicals in cat fur are unclear. BA activated several pus, lesions to this site result in significantly more time spent olfactory system structures and areas of the brain involved in in the open arms of an elevated plus maze (Kjelstrup et al., fi the neural processing of smell. Cat fur odor caused modest 2002). The ventral hippocampus also shows signi cant V1a activation in limbic structures compared to BA. We observed receptor mRNA expression (Ostrowski et al., 1994). We fi fi a signi cant effect of cat fur in anterior thalamic nucleus, observed signi cant effects of V1a blockade on cat but not BNST, ventral CA3, cortical amygdala, lateral amygdala, BA induced BOLD activation (Reed et al., 2013a, 2013b), which medial amygdala, olfactory tubercle, posterior amygdala, is consistent with some of the aforementioned studies. PAG, posterior hypothalamus, dorsomedial, dorsolateral and In sum, investigating the neural circuitry of unconditioned ventromedial regions of the striatum, nucleus accumbens fear and the modulatory role of OT and AVP may be improved core, subiculum, and ventral tegmental area (Reed et al., by the development of alternative models for eliciting signs of 2013a, 2013b). In spite of the shortfalls, an important finding fear that may be measurable from the rats. Something along was that OT and V1a receptors differentially modulated BA- the lines of what Ferris and colleagues reported for the and cat fur-induced BOLD signal responses in several brain resident-intruder model mentioned in the preceding section areas. BA-induced BOLD signal changes were curtailed in the (Ferris et al., 2008). The use of olfactory-mediated fear is cortical amygdala by the OT receptor antagonist (d(CH2)5-[Tyr complicated by the ambiguous outcomes of TMT (is it stress 2 4 9 (Me) ,Thr ,Tyr-NH2 ]Ornithine vasotocin). V1a receptor block- responsiveness or true innate anxiety/fear circuits) and the ade using [β-Mercapto-β,β-cyclopentamethylenepropionyl1,O- lack of detail on the chemical components of cat fur and feces me-Tyr2,Arg8]-Vasopressin reduced cat fur-induced BOLD that may trigger fear and defense reactions in rats. Moreover, signal changes in the amygdala, hippocampus and PAG. OT the use of centrally administered neuropeptides should be blockade had no effect on the response to cat fur in these improved by the administration of orally active agents as well regions. In the central amygdala both OT and V1a antagonists as using state of the art genetic tools to remotely drive activation curtailed the BOLD response to cat fur. of selective OT and AVP neurons in hypothalamic nuclei. Our The findings suggested that OT-mediated neurotransmis- first studies, however, indicate that OT and AVP may modulate sion is important in responding to a novel odor such as BA. fear and/or stress both centrally and during the initial processing

V1a-mediated neurotransmission appeared more closely stages in the olfactory system. This outcome is again consistent associated with cat fur induced BOLD activation across with the distribution of their corresponding receptors within several brain regions. This could indicate a closer association olfactory and limbic structures (Fig. 1). between V1a receptors and unconditioned fear than shown by OT receptors. However, the results only partly support this notion, and require improvements in the experimental 6. Potential role of V1a receptors in social model. The incorporation of conditioned fear stimuli, which neural processing in autism spectrum disorders holds significance to generalized anxiety and posttraumatic stress states, should provide the needed improvements. Original studies of AVP linked this peptide neuromodulator to It is noteworthy that we initially expected regions such as learning and memory functions (De Wied et al., 1975; De Wied the ventral BNST, amygdala, ventral hippocampus, lateral et al., 1993; Vawter et al., 1997). Although this was a large focus of septum, would distinguish the effects of cat odor from BA. classical studies it has subsided somewhat, but there is an 16 brain research 1580 (2014) 8–21

important neuroanatomical and neurophysiological association compared to controls were of greater magnitude. Differences between OT and AVP V1a/V1b receptor expression in the hippo- were observed in cortical and subcortical limbic regions, includ- campus and amygdala. Therefore, diseases in which the func- ing temporal, secondary motor and entorhinal cortices, basal tional activities of these peptides are dysregulated could lead to region of the amygdala, medial genticulate, posterior hypotha- impairments not only associated with anxiety, mood, violence, lamus and mammillary nuclei, and the substantia nigra pars but also can potentially underlie memory/learning impairments. reticulata. We also provided a primary sensory stimulus to study The neurophysiology underlying this possibility is studied very three brain regions of the visual pathway to examine whether scarcely. One such area involving complex deficits including differences in social neural processing could be attributable to social deficits, emotional and cognitive disturbances, is autism changes in primary sensory processing, albeit through a specific spectrum disorders (ASD). sensory system. Interestingly, we observed that the greater

A pilot study to examine the possible role of V1a receptors activation found with the social stimulus was not observed with was carried out in a well-studied animal model of ASD (Felix- the visual stimulus. Instead, the visual cortex of VPA rats showed Ortiz and Febo, 2012). ASD's, which includes autism, Asper- alowerpercentageBOLDsignalresponsesthatperhapsoccursas ger's syndrome, and pervasive developmental disorder not aresultoflowerneuralactivityinthisregion.TheeffectsofVPA otherwise specified (PD-NOS), are characterized by deficits in on primary sensory systems were thus distinct from subcortical verbal and non-verbal communication, reduced social inter- and higher order cortical processing of a socially salient stimulus. actions, and restricted range of interests and motor stereo- In order to examine whether VPA-induced effects on social typies. Genome-wide screening has identified numerous gene neural processing is associated with altered functionality mutations that might underlie ASD (Abrahams and activity through the V1a receptor, we included subgroups of

Geschwind, 2008). Among the candidate genes for ASD are animals that received an ICV injection of a V1a antagonist. A those encoding the V1a receptor (Kim et al., 2002), and the OT reduction in BOLD activity in these regions was observed with receptor (Abrahams and Geschwind, 2008), which are known V1a blockade. The antagonist on its own had no effect on the to modulate social behaviors in rodent models (Bielsky et al., pattern of brain activation shown in control animals. The 2004; Israel et al., 2008). effect of the antagonist was not found to affect primary

Genes encoding the V1a receptor have been linked to ASD sensory processing (visual). Therefore, the actions of the (Kim et al., 2002). The role of AVP in modulating the neural antagonist could be interpreted as occurring in conjunction response to a social stimulus was assessed in adult rats exposed with VPA-induced deficits in higher order processing, but not prenatally to valproic acid (VPA). VPA is a and primary visual processing. , given during pregnancy to women suffering The role of V1a receptors in modulating socially-relevant from epilepsy. Fetal exposure at embryonic day E20-E24 can neural circuits was influenced by prenatal VPA exposure. This fi produce neural tube defects and ASD-like de cits. In rats, is consistent with the role of the V1a receptor subtype in teratologic effects similar to those reported in humans are social behaviors, and might point to V1a receptors as a observed when exposure occurs between E9.5-12.5 (Arndt et al., substrate affected by prenatal VPA treatment. Thus, in a 2005; Miyazaki et al., 2005; Rodier et al., 1996; Stodgell et al., 2006). teratologic model recapitulating severe ASD-like disturbances Reports indicate that VPA treated rats show early signs of in rodents, we find a pattern of social responding that is developmental abnormalities (Markram et al., 2008; Rodier partly rescued by V1a blockade. These results are encouraging et al., 1996; Schneider et al., 2001; Schneider and Przewlocki, and support recent clinical trials examining the efficacy and 2005)andtheseareconsistentwithmotorandcognitivebeha- safety of a selective and orally active AVP antagonist (RG7314) vioral features partially resembling ASD. Neurobiological changes in individuals with autism spectrum disorders (http://clinical have also been reported and (Arndt et al., 2005; Ingram et al., trials.gov/show/NCT01793441)(Nightingale, 2012). According 2000; Rinaldi et al., 2008; Rodier et al., 1996), which parallels to our data obtained in a rat model of VPA-associated social fi fi postmortem pathological ndings in autistics. Recent evidence de cits, blocking V1a receptors may potentially be effective at indicates that neonatal VPA exposure can influence AVP immu- ameliorating the social and anxiogenic deficits observed in nolabeling in several regions of the rodent brain such as the autistic individuals. However, more animal imaging studies anterior hypothalamic area and mediodorsal thalamus, and are needed to examine specific mechanisms of action. influences olfactory-based social behavior in rodents (Murray Further data using neurochemical techniques are needed to fi et al., 2011). We tested whether prenatally exposed rats show con rm whether AVP, or V1a receptors, or both, are affected. abnormal neurophysiological responses to social and non-social The role of V1a receptors, and also OT receptors, in the (visual) stimuli (Felix-Ortiz and Febo, 2012). regulation of emotion, cognition and social behavior merits The experiments were designed to track the developmental further investigation, within the context of the neurobiology progress of various social behaviors from the early postnatal of ASD. Our results show a greater BOLD response to a social period to adulthood. It was anticipated that gestational VPA stimulus in the VPA rat. The V1a receptor antagonist blocks would dramatically alter the way the specific brain regions of this. Therefore, the greater ‘sensitivity’ to social stimuli could the rat respond to a social stimulus. Our neuroimaging data in part be mediated by excitatory AVP neurotransmission. provided evidence of a lasting effect of prenatal VPA on social neural processing (Felix-Ortiz and Febo, 2012). VPA-exposed rats show significantly greater BOLD signal responses to a 7. Conclusions stimulus juvenile rat than saline control animals. Even though the observed behavioral changes suggested a deficit, Over the past decade or so, we have taken a novel and non- the differences in neural responses to a social stimulus in VPA invasive approach to studying the varied roles of OT and AVP brain research 1580 (2014) 8–21 17

in modulating maternal, aggression, fear and social brain references networks. The initial results are promising and call for furthering creative ways to examine vital emotional and motivation states. Our neuroimaging findings skim the sur- Abrahams, B.S., Geschwind, D.H., 2008. Advances in autism face of the underlying cellular and molecular mechanisms genetics: on the threshold of a new neurobiology. Nat. Rev. and really provide what might be considered as ‘integrated Genet. 9, 341–355. network states'. That is, maps of how the brain under the Altemus, M., Cizza, G., Gold, P.W., 1992. Chronic fluoxetine treatment reduces hypothalamic vasopressin secretion tested conditions, initiates an effectively patterned response in vitro. Brain. Res. 593, 311–313. involving not just one region, but multiple disparate regions Appenrodt, E., Schnabel, R., Schwarzberg, H., 1998. Vasopressin across the brain. The integration of the multiple (otherwise administration modulates anxiety-related behavior in rats. unassociated) brain regions in these states is perhaps an area Physiol. Behav. 64, 543–547. that deserves deeper consideration. A resetting or synchro- Arndt, T.L., Stodgell, C.J., Rodier, P.M., 2005. The teratology of nization of population firing across distributed brain regions autism. Int. J. Dev. Neurosci. 23, 189–199. may underlie the observed putative network states under the Baldwin, B.A., Shillito, E.E., 1974. The effects of ablation of the olfactory bulbs on parturition and maternal behaviour in Soay influence of the triggering sensory stimuli selected for each sheep. Anim. Behav. 22, 220–223. study. fMRI in such a case is ideally suited for reliably Bale, T.L., Davis, A.M., Auger, A.P., Dorsa, D.M., McCarthy, M.M., examining the patterned neural responses in rats. 2001. CNS region-specific oxytocin receptor expression: We find that across the integrated network states (which importance in regulation of anxiety and sex behavior. involve many regions that have been associated indirectly, or J. Neurosci. 21, 2546–2552. casually, to cognitive, emotional and motivational mechan- Barberis, C., Tribollet, E., 1996. Vasopressin and oxytocin isms), OT and AVP exhibit modulatory functions. The obser- receptors in the central nervous system. Crit. Rev. Neurobiol. 10, 119–154. vation of these modulatory functions indirectly link these Bielsky, I.F., Hu, S.B., Szegda, K.L., Westphal, H., Young, L.J., 2004. neuropeptides to the brain networks and may be a rationale Profound impairment in social recognition and reduction in for targeting these substrates when the ‘network states' fail, anxiety-like behavior in vasopressin V1a receptor knockout as in psychiatric illnesses. Under normal conditions, we mice. Neuropsychopharmacology 29, 483–493. observe that OT is released within the lactating rat brain Blanchard, R.J., Blanchard, D.C., 1977. Aggressive behavior in the during suckling stimulation and therefore activates particular rat. Physiol. Behav. 1, 197–224. networks of regions in the maternal brain (Febo et al., 2005). Blume, A., Bosch, O.J., Miklos, S., Torner, L., Wales, L., Waldherr, M., Neumann, I.D., 2008. Oxytocin reduces anxiety via ERK1/2 AVP and OT play major roles in modulating brain regions activation: local effect within the rat hypothalamic involved emotional states (Reed et al., 2013a, 2013b) and paraventricular nucleus. Eur. J Neurosci. 27, 1947–1956. agonistic behaviors (Ferris et al., 2008). Across several of Bosch, O.J., Meddle, S.L., Beiderbeck, D.I., Douglas, A.J., Neumann, these studies, there are relatively consistent networks of I.D., 2005. Brain oxytocin correlates with maternal aggression: brain regions that arise within the context of their role in link to anxiety. J Neurosci. 25, 6807–6815. internal motivational and emotional states that are modu- Bosch, O.J., Neumann, I.D., 2008. Brain vasopressin is an lated by OT and AVP and their receptors. Most of these are important regulator of maternal behavior independent of dams’ trait anxiety. Proc. Natl. Acad. Sci. USA 105, summarized in Table 1 and Fig. 1. The brain regions include 17139–17144. as center points nuclei in the basal forebrain, amygdalar Bosch, O.J., Neumann, I.D., 2010. Vasopressin released within nuclei, midbrain regions and also include subregions of the the central amygdala promotes maternal aggression. Eur. cortex that were not previously associated with aggression, J. Neurosci. 31, 883–891. maternal care, and unconditioned fear. The results also point Bosch, O.J., Neumann, I.D., 2012. Both oxytocin and vasopressin to possible differences according to reproductive status. are mediators of maternal care and aggression in rodents: Although this was not directly examined in the above-cited from central release to sites of action. Hormones Behav. 61, 293–303. studies, there are clear differences between the studies in Bridges, R.S., Mann, P.E., 1994. Prolactin-brain interactions in the male rodents versus those including lactating rats. This induction of maternal behavior in rat. Psychoneuroendocrinology warrants further investigation (Carter, 2007), as it might be 19, 611–622. important from a drug discovery and treatment development Caffrey, M.K., Nephew, B.C., Febo, M., 2010. Central vasopressin standpoint. V1a receptors modulate neural processing in mothers facing intruder threat to pups. Neuropharmacology 58, 107–116. Carter, C.S., 2007. Sex differences in oxytocin and vasopressin: implications for autism spectrum disorders?. Behav Brain Res. 176, 170–186. Acknowledgments Carter, C.S., Grippo, A.J., Pournajafi-Nazarloo, H., Ruscio, M.G., Porges, S.W., 2008. Oxytocin, vasopressin and sociality. Prog. Support was provided in part by NIH grant DA019946. Its Brain. Res. 170, 331–336. contents are solely the responsibility of the authors and do Champagne, F., Diorio, J., Sharma, S., Meaney, M.J., 2001. not necessarily represent the official views of the National Naturally occurring variations in maternal behavior in the rat are associated with differences in estrogen-inducible Institute of Health. The NIH and NIDA had no role in study central oxytocin receptors. Proc. Natl. Acad. Sci. USA 98, design; in the collection, analysis and interpretation of data; 12736–12741. in the writing of the report; and in the decision to submit the Chen, W., Shields, J., Huang, W., King, J.A., 2009. Female fear: article for publication. The authors have no financial, com- influence of estrus cycle on behavioral response and neuronal mercial or personal conflict of interest to report. activation. Behav Brain Res. 201, 8–13. 18 brain research 1580 (2014) 8–21

Coccaro, E.F., Kavoussi, R.J., Hauger, R.L., Cooper, T.B., Ferris, C.F., terminalis is critical for fear behavior induced by 1998. Cerebrospinal fluid vasopressin levels: correlates with trimethylthiazoline, a component of fox odor. J. Neurosci. 25, aggression and serotonin function in personality-disordered 5998–6004. subjects. Arch. Gen. Psychiatry. 55, 708–714. Ferris, C., 2008. Functional magnetic resonance imaging and the Dantzer, R., Tazi, A., Bluthe, R.M., 1990. Cerebral lateralization of neurobiology of vasopressin and oxytocin. In: Neumann, I.D., olfactory-mediated affective processes in rats. Behav. Brain. Landgraf, R., (Eds.), Progress in Brain Research. Vol. 170, Res. 40, 53–60. pp. 305–319. Day, H.E., Masini, C.V., Campeau, S., 2004. The pattern of brain c- Ferris, C.F., Potegal, M., 1988. Vasopressin receptor blockade in the fos mRNA induced by a component of fox odor, 2,5-dihydro- anterior hypothalamus suppresses aggression in hamsters. 2,4,5-trimethylthiazoline (TMT), in rats, suggests both Physiol. Behav. 44, 235–239. systemic and processive stress characteristics. Brain. Res. Ferris, C.F., Pilapil, C.G., Hayden-Hixson, D., Wiley, R., Koh, E.T., 1025, 139–151. 1991. Evidence for two functionally and anatomically distinct De Kloet, E.R., Rotteveel, F., Voorhuis, T.A., Terlou, M., 1985. populations of magnocellular neurons in the golden hamster. Topography of binding sites for neurohypophyseal hormones J. Neuroendocrinol. 4, 193–205. in rat brain. Eur. J. Pharmacol. 110, 113–119. Ferris, C.F., 1996. Serotonin inhibits vasopressin facilitated De Wied, D., Bohus, B., Van Wimersma, B., T.B., 1975. Memory aggression in the Syrian hamster. In: Ferris, C., Grisso, T. deficit in rats with hereditary diabetes insipidus. Brain. Res. (Eds.), Understanding aggressive. Behavior in Children, vol. 85, 152–156. 794. New York Academy of Sciences, New York, pp. 98–103. de Wied, D., Gaffori, O., van Ree, J.M., de Jong, W., 1984. Central Ferris, C.F., Melloni Jr., R.H., Koppel, G., Perry, K.W., Fuller, R.W., target for the behavioural effects of vasopressin Delville, Y., 1997. Vasopressin/serotonin interactions in the neuropeptides. Nature 308, 276–278. anterior hypothalamus control aggressive behavior in golden De Wied, D., Diamant, M., Fodor, M., 1993. Central nervous hamsters. J. Neurosci. 17, 4331–4340. system effects of the neurohypophyseal hormones and Ferris, C.F., 2005. Vasopressin/oxytocin and aggression. In: Bock, related peptides. Front. Neuroendocrinol. 14, 251–302. G., Goode, J. (Eds.), Molecular Mechanisms Influencing Delville, Y., Mansour, K.M., Ferris, C.F., 1996. Serotonin blocks Aggressive Behaviours. Novartis Foundation Symposium, Vol. vasopressin-facilitated offensive aggression: interactions 268. Wiley, Chichester, UK, pp. 190–200. within the ventrolateral hypothalamus of golden hamsters. Ferris, C.F., Lu, S.-F., Messenger, T., Guillon, C.D., Koppel, G.A., Physiol. Behav. 59, 813–816. Miller, M.J., Heindel, N.D., Simon, N.G., 2006. Orally active Delville, Y., De Vries, G.J., Ferris, C.F., 2000. Neural connections of vasopressin V1a receptor antagonist, SRX251, selectively the anterior hypothalamus and agonistic behavior in golden blocks aggressive behavior. Pharm. Biochem. Behav. 83, hamsters. Brain. Behav. Evol. 55, 53–76. 169–174. Deschamps, S., Woodside, B., Walker, C.D., 2003. Pups presence Ferris, C.F., Stolberg, T., Kulkarni, P., Murugavel, M., Blanchard, R., eliminates the stress hyporesponsiveness of early lactating Blanchard, D.C., Febo, M., Brevard, M., Simon, a.N.G., 2008. females to a representing a threat to the Imaging the neural circuitry and chemical control of pups. J. Neuroendocrinol. 15, 486–497. aggressive motivation. BMC. Neurosci., 9. Devinsky, O., Morrell, M.J., Vogt, B.A., 1995. Contributions of anterior Fleming, A.S., Rosenblatt, J.S., 1974. Olfactory regulation of cingulate cortex to behaviour. Brain 118, 279–306 (Pt 1). maternal behavior in rats. I. Effects of olfactory bulb removal Dluzen, D.E., Muraoka, S., Engelmann, M., Landgraf, R., 1998. The in experienced and inexperienced lactating and cycling effects of infusion of arginine vasopressin, oxytocin, or their females. J. Comp. Physiol. Psychol. 86, 221–232. antagonists into the olfactory bulb upon social recognition Francis, D.D., Champagne, F.C., Meaney, M.J., 2000. Variations in responses in male rats. Peptides 19, 999–1005. maternal behaviour are associated with differences in Endres, T., Fendt, M., 2008. Inactivation of the lateral septum oxytocin receptor levels in the rat. J. Neuroendocrinol. 12, blocks fox odor-induced fear behavior. Neuroreport 19, 1145–1148. 667–670. Freund-Mercier, M.J., Stoeckel, M.E., Palacios, J.M., Pazos, A., Erskine, M.S., Barfield, R.J., Goldman, B.D., 1978. Intraspecific Reichhart, J.M., Porte, A., Richard, P., 1987. Pharmacological fighting during late pregnancy and lactation in rats and effects characteristics and anatomical distribution of [3H]oxytocin- of litter removal. Behav. Biol. 23, 206–218. binding sites in the Wistar rat brain studied by Febo, M., Numan, M., Ferris, C.F., 2005. Functional magnetic autoradiography. Neuroscience 20, 599–614. resonance imaging shows oxytocin activates brain regions Gabbott, P.L., Warner, T.A., Jays, P.R., Salway, P., Busby, S.J., 2005. associated with mother-pup bonding during suckling. Prefrontal cortex in the rat: projections to subcortical J. Neurosci. 25, 11637–11644. autonomic, motor, and limbic centers. J. Comp. Neurol. 492, Febo, M., Shields, J., Ferris, C.F., King, J.A., 2009. Oxytocin 145–177. modulates unconditioned fear response in lactating dams: an Gandelman, R., Zarrow, M.X., Denenberg, V.H., Myers, M., 1971. fMRI study. Brain. Res. 1302, 183–193. Olfactory bulb removal eliminates maternal behavior in the Ferris, C.F., Kulkarni, P., Sullivan Jr., J.M., Harder, J.A., Messenger, mouse. Science 171, 210–211. T.L., Febo, M., 2005. Pup suckling is more rewarding than Georges, F., Aston-Jones, G., 2001. Potent regulation of midbrain cocaine: evidence from functional magnetic resonance dopamine neurons by the bed nucleus of the stria terminalis. imaging and three- dimensional computational analysis. J. J Neurosci. 21, RC160. Neurosc. 25 (1), 149–156. Georges, F., Aston-Jones, G., 2002. Activation of ventral tegmental Felix-Ortiz, A.C., Febo, M., 2012. Gestational valproate alters bold area cells by the bed nucleus of the stria terminalis: a novel activation in response to complex social and primary sensory excitatory amino acid input to midbrain dopamine neurons. stimuli. PLoS. One 7, e37313. J. Neurosci. 22, 5173–5187. Fendt, M., Endres, T., Apfelbach, R., 2003. Temporary inactivation Gimpl, G., Fahrenholz, F., 2001. The oxytocin receptor system: of the bed nucleus of the stria terminalis but not of the Structure, function, and regulation. Physiol. Rev. 81, amygdala blocks freezing induced by trimethylthiazoline, a 629–683. component of fox feces. J. Neurosci. 23, 23–28. Goodson, J.L., Bass, A.H., 2001. Social behavior functions and Fendt, M., Siegl, S., Steiniger-Brach, B., 2005. Noradrenaline related anatomical characteristics of vasotocin/vasopressin transmission within the ventral bed nucleus of the stria system in vertebrates. Brain. Res. Rev. 35, 246–265. brain research 1580 (2014) 8–21 19

Grimes, J.M., Melloni Jr, R.H., 2002. Serotonin modulates offensive Larrazolo-Lopez, A., Kendrick, K.M., Aburto-Arciniega, M., attck in adolescent -treated hamsters. Pharm. Arriaga-Avila, V., Morimoto, S., Frias, M., Guevara-Guzman, R., Biochem. Behav. 73, 713–721. 2008. Vaginocervical stimulation enhances social recognition Grimes, J.M., Melloni Jr., R.H., 2005. Serotonin-1B receptor activity memory in rats via oxytocin release in the olfactory bulb. and expression modulate the aggression-stimulating effects Neuroscience 152, 585–593. of adolescent anabolic steroid exposure in hamsters. Behav. Levy, F., Kendrick, K.M., Goode, J.A., Guevara-Guzman, R., Neurosci. 119, 1184–1194. Keverne, E.B., 1995. Oxytocin and vasopressin release in the Guillon, C.D., Koppel, G.A., Brownstein, M.J., Chaney, M.O., Ferris, olfactory bulb of parturient ewes: changes with maternal C.F., Lu, S.-F., Fabio, K.M., Miller, M.J., Heindel, N.D., Hunden, D. experience and effects on acetylcholine, gamma- C., Cooper, R.D.G., Kaldor, S.W., Skelton, J.J., Dressman, B.A., aminobutyric acid, glutamate and noradrenaline release. Clay, M.P., Steinberg, M.I., Bruns, R.F., Simon, N.G., 2006. Brain. Res. 669, 197–206. Azetidinones as vasopressin V1a antagonists. Bioorg. Med. Lim, M.M., Murphy, A.Z., Young, L.J., 2004. Ventral striatopallidal Chem. 15, 2054–2080. oxytocin and vasopressin V1a receptors in the monogamous Heinrichs, M., Domes, G., Inga, D.N., Rainer, L., 2008. prairie vole (Microtus ochrogaster). J.Comp. Neurol. 468, 555–570. Neuropeptides and social behaviour: effects of oxytocin and Logothetis, N.K., 2003. The underpinnings of the BOLD functional vasopressin in humans. Progress in Brain. Research, Vol 170. magnetic resonance imaging signal. J. Neurosci. 23, 3963–3971. Elsevier337–350. Malin, E.L., McGaugh, J.L., 2006. Differential involvement of the Huber, D., Veinante, P., Stoop, R., 2005. Vasopressin and oxytocin hippocampus, anterior cingulate cortex, and basolateral excite distinct neuronal populations in the central amygdala. amygdala in memory for context and footshock. Proc. Natl. Science 308, 245–248. Acad. Sci USA 103, 1959–1963. Ingram, J.L., Peckham, S.M., Tisdale, B., Rodier, P.M., 2000. Prenatal Malin, E.L., Ibrahim, D.Y., Tu, J.W., McGaugh, J.L., 2007. Involvement of exposure of rats to valproic acid reproduces the cerebellar the rostral anterior cingulate cortex in consolidation of inhibitory anomalies associated with autism. Neurotoxicol. Teratol. 22, avoidance memory: interaction with the basolateral amygdala. 319–324. Neurobiol. Learn. Mem. 87, 295–302. Israel, S., Lerer, E., Shalev, I., Uzefovsky, F., Reibold, M., Bachner- Mann, P.E., Babb, J.A., 2004. Disinhibition of maternal behavior Melman, R., Granot, R., Bornstein, G., Knafo, A., Yirmiya, N., following neurotoxic lesions of the hypothalamus in Ebstein, R.P., 2008. Molecular genetic studies of the arginine primigravid rats. Brain. Res. 1025, 51–58. vasopressin 1a receptor (AVPR1a) and the oxytocin receptor Markram, K., Rinaldi, T., La Mendola, D., Sandi, C., Markram, H., 2008. (OXTR) in human behaviour: from autism to altruism with Abnormal fear conditioning and amygdala processing in an some notes in between. Prog. Brain. Res. 170, 435–449. animal model of autism. Neuropsychopharmacology 33, 901–912. Kendrick, K.M., Keverne, E.B., Chapman, C., Baldwin, B.A., 1988a. Miyazaki, K., Narita, N., Narita, M., 2005. Maternal administration Microdialysis measurement of oxytocin, aspartate, gamma- of thalidomide or valproic acid causes abnormal serotonergic aminobutyric acid and glutamate release from the olfactory neurons in the offspring: implication for pathogenesis of bulb of the sheep during vaginocervical stimulation. Brain. autism. Int. J. Dev. Neurosci. 23, 287–297. Res. 442, 171–174. Morrow, B.A., Redmond, A.J., Roth, R.H., Elsworth, J.D., 2000. The Kendrick, K.M., Keverne, E.B., Chapman, C., Baldwin, B.A., 1988b. predator odor, TMT, displays a unique, stress-like pattern of Intracranial dialysis measurement of oxytocin, monoamine dopaminergic and endocrinological activation in the rat. and uric acid release from the olfactory bulb and substantia Brain. Res. 864, 146–151. nigra of sheep during parturition, suckling, separation from lambs and eating. Brain. Res. 439, 1–10. Murgatroyd, C., Wigger, A., Frank, E., Singewald, N., Bunck, M., Kim, S.J., Young, L.J., Gonen, D., Veenstra-VanderWeele, J., Holsboer, F., Landgraf, R., Spengler, D., 2004. Impaired Courchesne, R., Courchesne, E., Lord, C., Leventhal, B.L., Cook repression at a vasopressin promoter polymorphism underlies Jr., E.H., Insel, T.R., 2002. Transmission disequilibrium testing overexpression of vasopressin in a rat model of trait anxiety. of arginine (AVPR1A) polymorphisms J. Neurosci. 24, 7762–7770. in autism. Mol. Psychiatry 7, 503–507. Murray, E.K., Varnum, M.M., Fernandez, J.L., de Vries, G.J., Forger, King, J.A., Garelick, T.S., Brevard, M.E., Chen, W., Messenger, T.L., N.G., 2011. Effects of neonatal treatment with valproic acid on Duong, T.Q., Ferris, C.F., 2005. Procedure for minimizing stress vasopressin immunoreactivity and olfactory behaviour in for fMRI studies in conscious rats. J. Neurosci. Methods 148, mice. J. Neuroendocrinol. 23, 906–914. 154–160. Nephew, B.C., Bridges, R.S., 2008. Arginine vasopressin V1a Kjelstrup, K.G., Tuvnes, F.A., Steffenach, H.A., Murison, R., Moser, receptor antagonist impairs maternal memory in rats. Physiol. E.I., Moser, M.B., 2002. Reduced fear expression after lesions of Behav. 95, 182–186. the ventral hippocampus. Proc Natl Acad Sci USA 99, Nephew, B.C., Caffrey, M.K., Felix-Ortiz, A.C., Ferris, C.F., Febo, M., 10825–10830. 2009. Blood oxygen level-dependent signal responses in Knobloch, H.S., Charlet, A., Hoffmann, L.C., Eliava, M., Khrulev, S., corticolimbic ‘emotions’ circuitry of lactating rats facing Cetin, A.H., Osten, P., Schwarz, M.K., Seeburg, P.H., Stoop, R., intruder threat to pups. Eur. J. Neurosci. 30, 934–945. Grinevich, V., 2012. Evoked axonal oxytocin release in the central Neumann, I., Ludwig, M., Engelmann, M., Pittman, Q.J., Landgraf, amygdala attenuates fear response. Neuron 73, 553–566. R., 1993a. Simultaneous microdialysis in blood and brain: Knox, D., Brothers, H., Norman, G.J., Berntson, G.G., 2008. Nucleus oxytocin and vasopressin release in response to central and basalis magnocellularis and substantia innominata peripheral osmotic stimulation and suckling in the rat. corticopetal cholinergic lesions attenuate freezing induced by Neuroendocrinology 58, 637–645. predator odor. Behav. Neurosci. 122, 601–610. Neumann, I., Russell, J.A., Landgraf, R., 1993b. Oxytocin and Kolunie, J.M., Stern, J.M., Barfield, R.J., 1994. Maternal aggression vasopressin release within the supraoptic and paraventricular in rats: effects of visual or auditory deprivation of the mother nuclei of pregnant, parturient and lactating rats: a and dyadic pattern of ultrasonic vocalizations. Behav. Neural. microdialysis study. Neuroscience 53, 65–75. Biol. 62, 41–49. Neumann, I., Koehler, E., Landgraf, R., Summy-Long, J., 1994. An Kolunie, J.M., Stern, J.M., 1995. Maternal aggression in rats: effects oxytocin receptor antagonist infused into the supraoptic nucleus

of olfactory bulbectomy, ZnSO4-induced anosmia, and attenuates intranuclear and peripheral release of oxytocin during vomeronasal organ removal. Horm. Behav. 29, 492–518. suckling in conscious rats. Endocrinology 134, 141–148. 20 brain research 1580 (2014) 8–21

Nightingale, S., 2012. From the analyst’s couch: autism spectrum valproic acid’s teratogenic mechanism of action. Neurotoxicol. disorders. Nature. Rev. Drug. Discovery 11, 745–746. Teratol. 28, 617–624. Numan, M., Numan, M., 1996. A lesion and neuroanatomical Sullivan, G.M., Apergis, J., Bush, D.E., Johnson, L.R., Hou, M., tract-tracing analysis of the role of the bed nucleus of the stria Ledoux, J.E., 2004. Lesions in the bed nucleus of the stria terminalis in retrieval behavior and other aspects of maternal terminalis disrupt corticosterone and freezing responses responsiveness in rats. Dev. Psychobiol. 29, 23–51. elicited by a contextual but not by a specific cue-conditioned Ongur, D., Price, J.L., 2000. The organization of networks within fear stimulus. Neuroscience 128, 7–14. the orbital and medial prefrontal cortex of rats, monkeys and Szot, P., Bale, T.L., Dorsa, D.M., 1994. Distribution of messenger humans. Cereb. Cortex 10, 206–219. RNA for the vasopressin V1a receptor in the CNS of male and Ostrowski, N.L., Lolait, S.J., Young 3rd, W.S., 1994. Cellular female rats. Brain. Res. Mol. Brain. Res. 24, 1–10. localization of vasopressin V1a receptor messenger Tribollet, E., Barberis, C., Dreifuss, J.J., Jard, S., 1988a. ribonucleic acid in adult male rat brain, pineal, and brain Autoradiographic localization of vasopressin and oxytocin vasculature. Endocrinology 135, 1511–1528. binding sites in rat kidney. Kidney Int. 33, 959–965. Pedersen, C., Caldwell, J., Walker, C., Ayers, G., Mason, G., 1994. Tribollet, E., Barberis, C., Jard, S., Dubios-Dauphin, M., Dreifuss, J. Oxytocin activates the postpartum onset of rat maternal J., 1988b. Localization and pharmacological characterization of behavior in the ventral tegmental and medial preoptic areas. high affinity binding sites for vasopressin and oxytocin in the Behav. Neurosci. 108, 1163–1171. rat brain by light microscopic autoradiography. Brain. Res. 442, Pedersen, C.A., Boccia, M.L., 2003. Oxytocin antagonism alters rat 105–118. dams’ oral grooming and upright posturing over pups. Physiol. Tribollet, E., Barberis, C., Jard, S., Dubois-Dauphin, M., Dreifuss, J.J., Behav. 80, 233–241. 1988c. Localization and pharmacological characterization of high Quignon, P., Giraud, M., Rimbault, M., Lavigne, P., Tacher, S., affinity binding sites for vasopressin and oxytocin in the rat brain Morin, E., Retout, E., Valin, A.S., Lindblad-Toh, K., Nicolas, J., by light microscopic autoradiography. Brain. Res. 442, 105–118. Galibert, F., 2005. The dog and rat olfactory receptor Vaccari, C., Lolait, S.J., Ostrowski, N.L., 1998. Comparative distribution repertoires. Genome Biol. 6, R83. of vasopressin V1b and oxytocin receptor messenger ribonucleic acids in brain. Endocrinology 139, 5015–5033. Reed, M.D., Pira, A.S., Febo, M., 2013a. Behavioral effects of acclimatization to restraint protocol used for awake animal van Leeuwen, F.W., van Heerikhuize, J., van der Meulen, G., imaging. J. Neurosci. Methods 217, 63–66. Wolters, P., 1985. Light microscopic autoradiographic localization of [3H]oxytocin binding sites in the rat brain, Reed, M.D., Price, K.E., Archbold, J., Moffa, A., Febo, M., 2013b. pituitary and mammary gland. Brain. Res. 359, 320–325. Predator odor-evoked BOLD activation in the awake rat: Vawter, M.P., De Wied, D., Van Ree, J.M., 1997. Vasopressin fragment modulation by oxytocin and V(1)a vasopressin receptor (4–8), improves long-term and short-term memory in the hole antagonists. Brain. Res. 1494, 70–83. board search task. Neuropeptides 31, 489–494. Ricci, L.A., Rasakham, K., Grimes, J.M., Melloni Jr, R.H., 2006. Veinante, P., Freund-Mercier, M.J., 1997. Distribution of Serotonin-1A receptor activity and expression modulate oxytocin- and vasopressin-binding sites in the rat extended adolescent anabolic/androgenic steroid-induced aggression in amygdala: a histoautoradiographic study. J. Comp. Neurol. hamsters. Pharmacol. Biochem. Behav. 85, 1–11. 383, 305–325. Rinaldi, T., Silberberg, G., Markram, H., 2008. Hyperconnectivity of Vertes, R.P., 2004. Differential projections of the infralimbic and local neocortical microcircuitry induced by prenatal exposure prelimbic cortex in the rat. Synapse 51, 32–58. to valproic acid. Cereb. Cortex 18, 763–770. Vogt, B.A., Finch, D.M., Olson, C.R., 1992. Functional heterogeneity Rodier, P.M., Ingram, J.L., Tisdale, B., Nelson, S., Romano, J., 1996. in cingulate cortex: the anterior executive and posterior Embryological origin for autism: developmental anomalies of evaluative regions. Cereb. Cortex. 2, 435–443. the cranial nerve motor nuclei. J. Comp. Neurol. 370, 247–261. Waldherr, M., Neumann, I.D., 2007. Centrally released oxytocin Rosen, J.B., Adamec, R.E., Thompson, B.L., 2005. Expression of egr- mediates mating-induced anxiolysis in male rats. Proc. Natl. 1 (zif268) mRNA in select fear-related brain regions following Acad. Sci USA 104, 16681–16684. exposure to a predator. Behav. Brain. Res. 162, 279–288. Wallace, K.J., Rosen, J.B., 2001. Neurotoxic lesions of the lateral Sanchez-Andrade, G., Kendrick, K.M., 2009. The main olfactory nucleus of the amygdala decrease conditioned fear but not system and social learning in mammals. Behav. Brain. Res. unconditioned fear of a predator odor: comparison with 200, 323–335. electrolytic lesions. J. Neurosci. 21, 3619–3627. Schneider, T., Labuz, D., Przewlocki, R., 2001. Nociceptive changes Windle, R.J., Shanks, N., Lightman, S.L., Ingram, C.D., 1997a. in rats after prenatal exposure to valproic acid. Pol. J. Central oxytocin administration reduces stress-induced Pharmacol 53, 531–534. corticosterone release and anxiety behavior in rats. Schneider, T., Przewlocki, R., 2005. Behavioral alterations in rats Endocrinology 138, 2829–2834. prenatally exposed to valproic acid: animal model of autism. Windle, R.J., Wood, S., Shanks, N., Perks, P., Conde, G.L., da Costa, Neuropsychopharmacology 30, 80–89. A.P.C., Ingram, C.D., Lightman, S.L., 1997b. Endocrine and Shmuel, A., Augath, M., Oeltermann, A., Logothetis, N.K., 2006. behavioural responses to noise stress:comparison of virgin Negative functional MRI response correlates with decreases in and lactating female ratsduring non-disrupted maternal neuronal activity in monkey visual area V1. Nat. Neurosci. 9, activity. J. Neuroendocrinol. 9, 407–414. 569–577. Winslow, J.T., Hastings, N., Carter, C.S., Harbaugh, C.R., Insel, T.R., Simon, N.G., Cologer-Clifford, A., Lu, S.F., McKenna, S.E., Hu, S., 1993. A role for central vasopressin in pair bonding in 1998. Testosterone and its metabolites modulate 5HT1A abd monogamous prairie voles. Nature 365, 545–548. 5HT1B agonist effects on internale aggression. Neurosci. Wozniak, D.F., Stewart, G.R., Finger, S., Olney, J.W., 1989. Biobehav. Rev. 23, 325–336. Comparison of behavioral effects of nucleus basalis Staples, L.G., McGregor, I.S., Apfelbach, R., Hunt, G.E., 2008. Cat magnocellularis lesions and somatosensory cortex ablation in odor, but not trimethylthiazoline (fox odor), activates the rat. Neuroscience 32, 685–700. accessory olfactory and defense-related brain regions in rats. Yoshimura, R., Kiyama, H., Kimura, T., Araki, T., Maeno, H., Neuroscience 151, 937–947. Tanizawa, O., Tohyama, M., 1993. Localization of oxytocin Stodgell, C.J., Ingram, J.L., O’Bara, M., Tisdale, B.K., Nau, H., Rodier, receptor messenger ribonucleic acid in the rat brain. P.M., 2006. Induction of the homeotic gene Hoxa1 through Endocrinology 133, 1239–1246. brain research 1580 (2014) 8–21 21

Young, L.J., 1999. Oxytocin and vasopressin receptors and species Yu, G.Z., Kaba, H., Okutani, F., Takahashi, S., Higuchi, T., 1996. The typical social behaviors. Horm. Behav. 36, 212–221. olfactory bulb: a critical site of action for oxytocin in the Young, L.J., Toloczko, D., Insel, T.R., 1999. Localization of induction of maternal behaviour in the rat. Neuroscience 72,

vasopressin (V1a) receptor binding and mRNA in the rhesus 1083–1088. monkey brain. J. Neuroendocrinol. 11, 291–297. brain research 1580 (2014) 124–132

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Oxytocin and vasopressin support distinct configurations of social synchrony

Yael Apter-Levi, Orna Zagoory-Sharon, Ruth Feldmann

Department of Psychology, Gonda Brain Sciences Center, Bar-Ilan University, Ramat-Gan 52900, Israel article info abstract

Article history: Social synchrony – the coordination of behavior between interacting partners during social Accepted 27 October 2013 contact – is learned within the parent-infant bond and appears in a unique form in mothers Available online 1 November 2013 and fathers. In this study, we examined hormonal effects of OT and AVP on maternal and Keywords: paternal behavioral patterns and detail the processes of parent-infant social synchrony as Oxyctocin they combine with hormonal activity. Participants included 119 mothers and fathers (not – Vasopressin couples) and their 4 6 month-old infants. Baseline OT and AVP were collected from parents fi Parenting and a 10-minute face-to-face interaction with the infant was lmed. Interactions were micro- Synchrony coded for parent-child contact, social signals, and social- versus-object focused play. Propor- Social behavior tions and lag-sequential patterns of social behaviors were computed. Mothers provided more affectionate contact, while fathers provided more stimulatory contact. Parents with high OT levels displayed significantly more affectionate contact compared to parents with low OT and constructed the interaction towards readiness for social engagement by increasing social salience in response to infant social gaze. In contrast, parents with high AVP engaged in stimulatory contact and tended to increase object-salience when infants showed bids for social engagement. OT levels were independently predicted by the amount of affectionate contact and the durations of gaze synchrony, whereas AVP levels were predicted by stimulatory contact, joint attention to objects, and the parent increasing object salience following infant social gaze. Results further specify how synchronous bio-behavioral processes with mother and father support the human infant's entry into the family unit and prepare the child for joining the larger social world. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2013 Elsevier B.V. All rights reserved.

1. Introduction of adequate caregiving and the formation of the parent-infant bond (Curley and Keverne, 2005). Thus, pregnancy and child- The nature of the affiliative bond between infants and their birth occur in the context of marked changes in maternal and parents is crucial for understanding human relationships paternal brain areas implicated in motivation, nurturance, and the developmental psychopathologies that result from and attention (Atzil et al., 2011; Kinsley and Amory-Meyer, its malfunction (Douglas, 2010). Across mammalian species, 2011; Mosek-Eilon et al., 2013; Swain et al., 2007). Mothers' the transition to parenthood involves a major neuro- and fathers' brains undergo changes and become sensitive to hormonal reorganization that is essential for the provision their infants' cues (Kim et al., 2010), and similar changes are

nCorresponding author. Fax: 972 3 535 0267. þ E-mail address: [email protected] (R. Feldman).

0006-8993/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.brainres.2013.10.052 brain research 1580 (2014) 124–132 125

observed in hormonal systems (Feldman, 2012). These neuro- has shown that AVP promotes social recognition in both hormonal changes support the expression of the species- animals, especially rodents, (Caldwell et al., 2008), and typical behavioral repertoire in mothers and fathers that human males (Guastella et al., 2010). Regions characterized prompt the parent-infant affiliative attachment (Feldman as part of the AVP circuitry are implicated in socio-cognitive et al., 2007). A basic concept in the understanding of this processes in both humans and rodents (Goodson and affiliative bond is that of bio-behavioral synchrony (Feldman, Thompson, 2010). This AVP-brain associations may represent 2007, 2012; Feldman et al., 2012). Human studies have shown elevated AVP-dependent vigilance, which supports the that parent-infant social synchrony – the coordination of father's ability to read the intention of others in order to parental behavior with the infant's social signals – describes a defend mother and young (Atzil et al., 2012; Thompson et al., distinct and stable behavior constellation that is uniquely 2006). In contrast, in women, AVP was found to support the expressed in mothers and fathers (Feldman, 2007, 2012; mother's ability to befriend with others. Thus, AVP may Feldman et al., 2012). The human maternal behavioral repertoire prompt differential social strategies in social contexts in is largely based on eye contact with the infant, “motherese” women and men (Thompson et al., 2006). vocalizations, affectionate touch, and the appropriate and OT effects on human social functioning, however, are not synchronous adaptation of these behaviors to infant respon- uniform and depend on the individual's attachment history siveness (Feldman and Eidelman, 2004, 2007). The human and social skills (Bartz and Hollander, 2006; Weisman et al., paternal constellation, on the other hand, involves redirec- 2013b). The influence of OT on social emotion processing in tion of infant attention to the environment, stimulatory humans appears to depend, at least in part, on gender (Gamer contact, and joint attention in the exploration of objects et al., 2010), and OT significantly increased activations in (Feldman et al., 2013; Feldman et al., 2013; Parke, 1996), brain areas involved with emotion encoding and empathy in although mothers and fathers both employ the “maternal” females and not in males (Decety, 2010). This may imply that and “paternal” repertoire during social play. These behaviors OT influences prompts different parental behaviors in parallel those described in the classical animal literature, mothers and fathers. Interestingly, in mothers, but not in mainly in rodents, that link maternal behavior to motiva- fathers, plasma OT correlated with limbic activations (Zink tional and affective neural systems (Champagne et al., 2008). et al., 2011). It thus appears that maternal instinctual care There is currently much support for the notion that may originate from a limbic OT-sensitive motivational circuit, human attachment develops within the matrix of biological while fathering is acquired by experience, influenced by attunement and close behavioral synchrony (Atzil et al., 2013; social-cognitive processes and AVP. Feldman, 2012; Feldman et al., 2011, 2012). An important area Brain activity–OT correlations provide additional support to of research in the neurobiology of attachment has been the the notion that mothering is guided by greater motivational- hormonal system, in particular the posterior hypophysial emotional focus whereas fathering by a more socio-cognitive peptides oxytocin (OT) and vasopressin (AVP). Accumulating executive focus. We found that plasma OT levels may evidence has shown in both animals and humans that these reflect enhanced maternal but not paternal brain activity in hormones are indispensable elements in the developing limbic-emotional brain areas. In contrast, father OT corre- formation of relationships, affecting individual differences lated with higher activations in socio-cognitive circuits, in parenting behavior, social recognition, and affiliative whereas AVP was linked with fathers' amygdala activations behaviors (for review, Feldman, 2012; Ishak et al., 2011; (Atzil et al., 2012). It has further been reported that when Skuse and Gallagher, 2011). Studies in rodents (prairie voles) fathers received exogenous OT, their infants' showed a indicate that variations in maternal behavior based on dis- comparative increased levels of salivary OT and both partners tinct patterns of mothering correlate with a specific bio- engaged in greater toy exploration (Weisman et al., 2012). behavioral profile and greater OT receptor densities in both Furthermore, infant OT response correlated with the beha- mother and child (Olazábal and Young, 2006). In humans, OT vioral repertoire typical of the father–infant bond, including levels in parent and child are inter-related and depend on the paternal stimulation and joint object exploration (Feldman degree of interactive synchrony (Feldman et al., 2010a, 2011), et al., 2010b). Similarly, Naber et al. (2010), showed that OT including gaze synchrony and the matching of affective administration increased fathers' stimulatory and explora- expression. Mothers who engaged in more synchronous tory play with their toddlers. interactions showed more coherent activations of the amyg- The aforementioned synchronous bio-behavioral pro- dala and nucleus accumbens (NAcc) to their infant's stimuli, cesses allow the human infant to enter the social world and these activations correlated with maternal plasma OT of the family unit and to prepare for joining the larger (Atzil et al., 2011). Parallel to research in mice pointing to social group. The bio-behavioral synchrony conceptual model associations between mothers' and fathers' physiological and (Feldman, 2012; Feldman et al., 2012) postulates that the behavioral responses in the context of infant cues (Franssen formation of human attachment includes a finely-tuned et al., 2011), we found correlations between OT levels in adaptation of the parent and infant's neural function. Still, human mothers, fathers, and infants (Feldman et al., 2013) there are some major gaps in the literature that require as well as synchrony between mothers' and fathers' brain attention. The vast majority of studies have focused on response to their own infant's stimuli (Atzil et al., 2012). maternal behavior and there is a relative lack of studies on In contrast to OT, very little is known about the effects of fathers. Furthermore, most studies have not differentiated AVP on human parenting. In rodents, AVP has been asso- between the effects of the two “bonding” hormones, OT and ciated with male bonding and defensive and territorial AVP, on patterns of parental–infant bonding. Research on behavior in rodents (Bielsky et al., 2005), and recent research AVP is predominantly male oriented as AVP has been mostly 126 brain research 1580 (2014) 124–132

studied in the context of autism and aggression. There is a paucity of AVP research in humans, with no prior study 3. Results testing the links between AVP and parenting behavior. Finally, there is a lack of understanding of how OT and AVP 3.1. Differences in OT, AVP, and social behavior between differentially structure the behavioral repertoire involved in mothers and fathers social interactive processes. fi As such, the goals of the present study were to compare As a rst step, we examined whether plasma OT and AVP levels the hormonal effects of OT and AVP on both maternal and differ between mothers and father. No differences were found fi fi paternal parenting behaviors and to detail the processes of in baseline OT and AVP, con rming our previous ndings (Atzil parent-infant interaction as they synchronize with hormonal et al., 2012; Feldman et al., 2010a, 2011). Levels of OT for mothers were 388.05 (SD 205.95) and for fathers 391.18 activity. ¼ (SD 159.72), F (1, 118) .010, NS. Levels of AVP for mothers Our major hypothesis was that both OT and AVP levels ¼ ¼ were 223.29 (SD 85.93) and for fathers 201.37 (SD 86.37), would be associated with the social behaviors that form the ¼ ¼ F(1,118) 1.855, NS. OT and AVP were each divided into high basis of parental care. We also hypothesized that hormonal ¼ fi and low groups based on the median split (OT median 332.8 levels would correlate with the speci c ways parents syn- ¼ pM, AVP median 200.2 pM). Consistent with our hypothesis, chronize and structure social interaction with their offspring ¼ parents' OT and AVP were inter-related, r .21, p .024. and that this would differ between OT and AVP. Finally, based ¼ ¼ on research showing that OT administration increases Three MANOVAS examined differences between mothers peripheral AVP (Weisman et al., 2013a), suggesting inter- and fathers in (a) proportions of touch patterns (affectionate relatedness between the two hormones, we expected correla- touch, stimulatory touch), (b) synchrony variables (social gaze tions between maternal and paternal OT and AVP. synchrony, joint attention,), and (c) lag-sequential patterns 119 parents (not couples) and their 4–6 month-old infants (Infant gaze at parent–parent increase social salience: Infant gaze participated in the study. Parental plasma OT and AVP levels at parent–parent increase object salience: Infant gaze at parent– were measured. Parent infant interaction was coded system- parent affectionate touch, and Infant gaze at parent–parent stimu- atically from video recordings focusing on Affectionate and latory touch). Stimulatory Contact, Joint Attention, and Gaze Synchrony. In MANOVA for proportions of touch showed an overall effect for parent gender; Wilks' F (2, 116) 35.24, po.000, addition behavioral sequences of the parent's reaction to the ¼ Effect Size (ES) .37. Mothers provided significantly more infant's gaze, parent increase social or object salience were ¼ affectionate touch than fathers, Univariate F (1, 116) 19.80, assessed. ¼ po.000, ES .145, while fathers exhibited substantially more ¼ stimulatory touch, F (1116) 47.78, po.000, ES .29. ¼ ¼ 2. Statistical analysis MANOVA for the synchrony variables showed no effect of parent gender, Wilks' F (2, 116) 1.02, p .38, NS. Similarly, ¼ ¼ In the first section, ANOVAs examined differences between MANOVA for lag-sequential patterns showed no parent gender effect, Wilks' F (4, 114) .49, p .73, NS. These findings maternal and paternal OT and AVP. Next, three MANOVAs ¼ ¼ examined differences between mothers and fathers in: suggest that while mothers and fathers provide different (a) touch behavior; (b) synchrony variables, and (c) lag- types of touch, the sequential organization of their social sequential patterns. In the second section, similar three behavior during play with their infant does not differ. MANOVAs were computed twice: once for parents with high versus low OT and once for parents with high versus low 3.2. Differences in social behavior between parents of high AVP. In the third section, Pearson's correlations tested inter- and low OT and AVP relationships between hormones and behavior. Finally, two hierarchical regression models were computed predicting Next, similar three MANOVAs examined differences between parents' OT and AVP from touch behavior, synchrony variables, parents high and low in OT, and, following, the same three and lag-sequential patterns. MANOVAs tested those high and low in AVP.

Table 1 – Social Behavior in parents with high and low oxytocin.

Social Behavior High OT Low OT F

Mean SEM Mean SEM

Affectionate contact 56.82 2.6 42.93 3.07 11.86nn Stimulatory contact 27.88 1.77 28.38 1.52 .05 Gaze synchrony 6.75 .85 5.35 .76 1.51 Joint attention 3.51 .33 4.37 .4 2.72 Increasing social salience 1.14 .16 .73 .11 4.447n Increasing object salience .46 .11 .37 .09 .42

n po.05. nn po.01. brain research 1580 (2014) 124–132 127

3.2.1. High and low oxytocin Parental touch – MANOVA revealed a significant effect for Means and SD for all variables according to parental OT OT in parental touch; F (2, 116) 5.94, p .003. ES .093. ¼ ¼ ¼ appear in Tables 1 and 2. Univariate tests showed that parents with high OT provided

Table 2 – Social behavior in parents with high and low vasopressin.

High AVP Low AVP F Social Behavior Mean SEM Mean SEM

Affectionate contact 53.28 3.06 46.41 2.85 2.7 Stimulatory contact 30.5 1.53 25.77 1.7 4.3n Gaze synchrony 6.02 .83 6.97 .79 .002 Joint attention 3.5 .33 4.37 .4 2.72 Increasing social salience .96 .16 .92 .11 .03 Increasing object salience .56 .12 .27 .57 4.45n

Differences between parents' high and low OT levels in means and SEM of social measures in child–parent interaction. n po.05.

Proportion of Contact

60 ** High OT

55 Low OT 50

Proportions 45

40 Affectionate Contact

Lag Sequential 0.6 1.3 * 0.4 * 1.0

0.2 0.7 Frequency Frequency Frequency Frequency

0.0 0.4 Infant social gaze then parent affectionate infant social gaze then parent increases touch social salience

Fig. 1 – Parent–child contact and sequential patterns in parents with high and low oxytocin footer: *po.05, **po.01.

Proportion of Contact

35 * High AVP

30 Low AVP

25 Proportions 20 Stimulatory Contact

Lag Sequential 1.0 0.8 * 0.8 * 0.6

0.6 0.4

Frequency Frequency 0.4 0.2 Frequency Frequency

0.2 0.0 Infant gaze to parent then parent Infant social gaze then parent increases stimulatory touch obejct salience

Fig. 2 – Parent–child contact and sequential patterns in parents with high and low oxytocin footer: *po.05. 128 brain research 1580 (2014) 124–132

Table 3 – Correlations of hormones and social behavior.

Social Behavior OT AVP Affectionate Stimulatory Gaze Joint Increasing Increasing contact contact synchrony attention social object salience salience

Increasing object .06 .192n .094 .001 .159n .03 .15 1 À À $ salience Increasing social .191n 02 .004 .045 .008 .084 1 À À À salience Joint attention .04 15 .15nn 18n .03 1 À À Gaze synchrony .11 .004 .074 .04 1 À Stimulatory contact .02 .19n .163n 1 À À Affectionate contact .303n .15 1 AVP .04 1 À OT 1

n po.05. nn p .051. ¼

substantially more affectionate touch to their infants, 3.4. Predicting parental OT and AVP F (1116) 11.86, p .001, ES .092, and no differences emerged ¼ ¼ ¼ in stimulatory touch (Fig. 1). Finally, two regression equations were computed predicting Synchrony variables – no overall difference was found parents' OT and AVP from social behavior. Predictors were for OT. entered in seven theoretically-determined steps. Parent- Sequential patterns – MANOVA revealed a significant OT gender was entered in the first step to partial out gender effect for sequential patterns; F (2, 116) 2.685, p .042, ES affect. Following, the two touch patterns, two synchrony ¼ ¼ ¼ .082. Univariate tests showed thatinteractionsofparentswith variables, and two lag-sequential patterns were entered high OT included higher frequencies of the sequence: infant (Table 4). Both models were significant explaining 20% and social gaze then parent affectionate touch, F (1116) 4.43, 31% of the variability in OT and AVP respectively. ¼ p .037, ES .037, and infant social gaze then parent increase OT levels were independently predicted by the amount of ¼ ¼ social salience, F (1,116) 3.81, p .05, ES .033 (Fig. 1). affectionate contact, and the durations of gaze synchrony. ¼ ¼ ¼ AVP levels were uniquely predicted by the parent's stimula- tory contact, the duration of parent-infant joint attention to 3.2.2. High and low vasopressin object, and the parent responding to infant social gaze by Parental touch – MANOVA indicated significant main effect for increasing object salience. AVP, F (2, 116) 4.29, p .016. ES .069. Univariate tests ¼ ¼ ¼ showed that parents with high OT provided more stimulatory touch, F (1116) 4.29, p .04, ES .035, and no differences were ¼ ¼ ¼ 4. Discussion found for affectionate touch (Fig. 2). Synchrony variables – no overall effects were found for AVP. This study provides further evidence for the essential role of Sequential patterns – MANOVA revealed a significant AVP the pituitary hormones in parenting and for the complex and effect for sequential patterns; F (2, 116) 2.372, p .05. ES ¼ ¼ ¼ intricate relationship between the OT and AVP systems and .078. Univariate tests showed that parents with high AVP both mothering and fathering. To our knowledge, this is the tended to respond to infant social bids with increasing object first study to examine plasma AVP in human mothers and salience and stimulatory touch. Differences were for: infant fathers in comparison to plasma OT and parenting behavior. social gaze then parent stimulatory touch, F (1,116) 5.73, ¼ We found that each hormone is associated with both mater- p .018, ES .048, and infant social gaze then parent increase ¼ ¼ nal and paternal behavior but in distinct, sometimes over- object salience, F (1,116) 3.88, p .05, ES .033 (Fig. 2). ¼ ¼ ¼ lapping ways. It may thus be concluded that each hormone promotes a cascade of specific sequential behaviors which 3.3. Correlations between OT, AVP, and parental social form the basic elements of parenting. Specifically, our central behavior and synchronous sequential patterns hypothesis – that hormonal levels would correlate with the ways in which parents structure social interactions with their Pearson's correlations (Table 3) indicate that parental OT was offspring and that these correlations would differ for OT and associated with greater affectionate contact and greater AVP – was, in the main, supported by the findings. frequency of the sequence infant social gaze-parent increasing Not surprisingly, our behavioral observations showed social salience. Parents' AVP correlated with more stimulatory how affectionate contact was more characteristic of mothers, contact and longer durations of joint attention. Stimulatory while fathers showed more stimulatory contact. This is contact correlated with joint attention, and gaze synchrony consistent with studies in both rodents and humans (Szyf with less parental increasing object salience following infant et al., 2007; Feldman et al., 2010b). It thus appears that social gaze. paternal bonding and maternal bonding differ in their basic brain research 1580 (2014) 124–132 129

Table 4 – Predicting parent oxytocin (OT) and vasopressin (AVP).

Parental OT Parental AVP Social Behavior Beta R2 change F change Beta R2 change F change

Parent gender .07 .00 .43 .09 .02 1.85 À Affectionate contact .36nn .12 16.39nnn .07 .00 .96 Stimulatory contact .15 .02 2.57 .52nnn .21 22.92nnn Gaze synchrony .22n .05 5.81n .09 .00 .87 Joint attention .04 .00 .005 .20n .04 5.79nn Increasing social salience .03 .00 .31 .07 .00 1.39 Increasing object salience .08 .00 .98 .19n .03 3.41n R2 total .20, F (7, 110) 3.81, po.001 .31 F (7, 110) 7.13, po.001 ¼ ¼ ¼ n po.05. nn po.001. nnn po.0001.

behavioral structure. Importantly, irrespective of gender, high Research supports the importantroleofAVPinparenting.In levels of OT correlated with affectionate contact, response to humans, the lowest levels of maternal sensitivity were found infant gaze, and behaviors of high social salience. In contrast, among mothers who had a combination of high levels of early high AVP levels correlated with joint attention to inanimate adversity and a variant of AVPR1A gene (Bisceglia et al., 2012). objects and to behaviors with more object salience. Further- Fathers who had histories of adversity also showed hyper more, specific behaviors were shown to predict hormonal excretion of AVP that was related to differential activation of levels. Affectionate warm contact and duration of gaze syn- the amygdala, which in itself has shown to be a marker for chrony predicted OT levels, emphasizing that the parent's parental attunement to the infant's emotional state (Seifritz ability to react to their infants synchronously related to OT. et al., 2003). Similarly, AVP was found to be a component of Predictors of AVP, on the other hand, were more related to prairie vole paternal care (Ross et al., 2009)andinrats,males stimulatory contact, durations of joint attention, and the with a higher density of V1aRs in the lateral septum are more parent responding to infant social gaze by increasing object likely to provide paternal behavior. Furthermore, AVP has the salience in both mothers and fathers. These findings are capacity to bind not only to AVP receptors but also to the OXTR, supportive of those reported in previous research both in indicating that it has the potential to modulate the activity of bi-parental rodents (e.g. Woller et al., 2012; Veenema, 2012) various receptor subtypes (Kinsley and Amory-Meyer, 2011). and humans (e.g. Feldman et al., 2012; Atzil et al., 2012; Similar co-activation of the OTandAVPsystemshasrecently Gordon et al., 2010b, 2011; Skuse and Gallagher, 2011). The been shown in our lab, with the administration of OT reliably major novel aspects of this study, however, are the findings increasing salivary AVP levels across the first hour after that OT and AVP effects are important for both fathers and administration (Weisman et al., 2013a). However, apart from mothers and to the way they sequentially segment parental an earlier study by our group which showed a correlation behavior to direct infants to the social context or to features between social cognitive circuits in fathers and serum AVP, of the environment. the current study is the first to show a relationship between Fatherhood in itself can bring about hormonal changes in AVP and fathering in humans. the father, particularly when fathers become increasingly Oxytocin has been repeatedly implicated in the expression involved in parenting. In the context of evolving cultural of human maternal behavior with higher OT associated with and social values that underscore father involvement in sensitive and adaptive maternal care (Feldman et al., 2010; childcare (Lamb, 2010), understanding these processes in Gordon et al., 2010a, b; Levine et al., 2007), and the current the context of the bio-behavioral synchrony conceptual model results are consistent with these studies. Our findings extend becomes increasingly relevant (Feldman, 2012; Feldman et al., previous research by showing how OT is associated with 2012). Our results are consistent with perspectives that some of the more complex elements of human parenting. In a suggest a common neuroendocrine pathway in the develop- previous study we tested differences between OT and pro- ment of fathering and mothering in nonhuman primate's lactin in fathers and there is an interesting similarity in the biparental species and men (Wynne-Edwards, 2001). Whereas relative contribution of OT and AVP to the OT/PRL effects on in the past, maternal sensitivity has been stressed as vital for paternal behavior (Gordon et al., 2010c). Thus both AVP and infant development, sensitivity of both spouses is becoming PRL were associated with a specific aspect of paternal recognized as critical for optimal growth and longitudinal behavior not covered by OT: PRL with father facilitation of studies specify the unique contributions of sensitive mother- the child's exploratory behavior and OT with the father– ing and fathering to child social development (Feldman and infant affect Synchrony. In this context, AVP seems to be Masalha, 2010; Feldman et al., 2013). Our study provides a more associated with the cognitive-stimulatory aspects of the biological justification for stressing both maternal and pater- interaction. nal sensitive responsiveness and show that the link between It appears that early parent-child relationships take place OT and AVP with the unique constellations of parenting in the context of a network of bio-behavioral experiences that behavior appears in both maternal and paternal play. shape children's affiliative biology and social behavior across 130 brain research 1580 (2014) 124–132

multiple attachments (Feldman et al., 2013). Future research delivery, received an Apgar score of 9.40 (SD 1.56), and 55% ¼ will naturally expand our understanding of this network. It is were firstborns. Infants were healthy since birth, parents becoming increasingly evident that complex behaviors are were screened for depression and anxiety, and all parents rarely influenced by a single locus of main effect and are reported sharing childcare responsibilities. The study was subject to the influence of multiple neurohormonal systems approved by the Institutional Review Board and all parents and environmental conditions (Petronis, 2010). Possible areas signed an informed consent. for future research are the dopamine (DA) and serotonergic (5HT) system. Evidence is emerging implicating dopamine-OT 5.2. Procedure interactions in the modulation of neural circuits that influ- ence affiliative behaviors in rats (Shahrokh et al., 2010). In Parents and infants arrived at the lab during the early afternoon humans, receptor binding sites of OT and of DA tend to (1–4PM)andfollowinganacquaintingperiodplasmasamples coexist in several brain regions that are central for the were collected. Following, parentandchildenteredanobserva- expression of parental care (Skuse and Gallagher, 2011). There tion room with an infant-seat mounted on a table and were is also considerable evidence for the relations between OT filmed from an adjoined room by two cameras that were and the serotonin system. Animals exposed to elevated integrated into single frame using a split-screen generator. serotonin during early development have reduced OT expres- Parents were asked to engage in a fifteen-minute interaction sion and loss of OT-containing cells in the paraventricular that would include any type of touch they typically use. Parents nucleus in adulthood. This reduction is associated with were then interviewed and completed self-report measures. reduced maternal bonding and socially explorative behaviors found in rats (McNamara et al., 2008). As such, a more 5.3. Hormone collection and analysis comprehensive assessment of affiliative hormones within the context of multiple hormonal systems may shed further Blood was drawn from the antecubital-vein of the parents light on the neuroendocrine foundation of maternal and into 9 mL chilled vacutainer tubes containing lithium-heparin paternal care. that were supplemented with 400 KIU of Trasylol (Bayer, Leverkusen, Germany) per 1 mL blood. Samples were kept 4.1. Limitations ice-chilled for up to 2 h before centrifuged at 4 1C at 1000 g for 15 min. Supernatants were collected and stored at 70 1C À The major limitation of the study is that it is cross sectional and until assayed. Parents were asked to refrain from food intake thus cause–effect relationships cannot be inferred. As such, it is 30-minutes before blood draw. Maternal blood was drawn at unclear whether the hormonal changes result from the beha- least 30 min after nursing and 30 min before nursing. Pre- vioral changes or vice versa and prospective follow up studies vious studies (Feldman et al., 2007, 2010b, Gordon et al., are required to address this issue. In addition, a single study can 2010a) showed no differences between plasma OT levels in only include a limited number of hormones. It is clear that OT breastfeeding and non-breastfeeding mothers when OT is not and AVP do not act in a vacuum and are only part of multiple measured around breastfeeding. Determinations of hor- hormones and neurotransmitters which act in as a network to mones were performed using a commercial oxytocin and support parental care. As such, any linear model must neces- vasopressin enzyme-linked-immunosorbent-assay (ELISA) kit sarily be limited in its ability to explain complex human (Assay Design, Ann Arbor, Michigan) as described in ours and behavior such as parenting. Nonetheless, we believe that this others' research on plasma oxytocin and vasopressin (Carter study provides information that may lead to further program- et al., 2007; Feldman et al., 2007, 2010, 2011; Weisman et al., matic hypothesis-driven research and ultimately to a better 2013a). Measurements were performed in duplicate and the understanding of the issues involved in human bonding and concentrations of samples were calculated by MATLAB-7 parenting. Such understanding may, in turn, lead to the devel- according to relevant standard curves. The intra-assay and opment of therapies to help young mothers and fathers provide inter-assay coefficients for oxytocin were less than 7% and the carefully-synchronous parenting their infant requires for 15.8%, respectively. The intra-assay and inter-assay coefficients his or her physiological, social, and emotional growth. for Vasopressin were less than 3.9% and 16.9%, respectively.

5.4. Coding of parent-infant interaction 5. Experimental procedures Interactions were micro-coded on a computerized system in. 5.1. Participants 01-second frame using our well-validated micro-coding scheme (Atzil et al., 2012; Feldman and Eidelman, 2004, Participants were 119 parents and their infants, including 71 2007; Feldman et al., 2011; Feldman et al., 2010b), which has mothers and 48 fathers (not couples) and their 4–6 month-old shown associations with multiple parental hormones and infants (M 167.4 days, SD 12.3). Parents were of middle- patterns of brain activations. Four categories of parent and ¼ ¼ class SES, healthy, and with at least 12 years of education. infant's behavior were: Gaze, Affect, Vocalization, and Touch, Mothers' age was, M 28.9, SD 5.22 years, and education, each containing a set of mutually-exclusive codes. Codes ¼ ¼ M 15.17, SD 2.47 years, and 81.3% were breastfeeding. included: Gaze: social gaze to partner, gaze to object (parent or ¼ ¼ Fathers' age was, M 29.3, SD 4.26 years and education, child looking at object other than the focus of the partner's ¼ ¼ M 15.53, SD 2.71 years. Infants were born at term (birth- gaze), joint attention (parent and child looking at same ¼ ¼ weight: M 3319.4 gr. SD 452.1), mainly (96.3%) by vaginal object), gaze aversion. Affect: parent – positive, neutral, ¼ ¼ brain research 1580 (2014) 124–132 131

negative-withdrawn, negative-angry; infant- positive neutral, Bartz, J.A., Hollander, E., 2006. The neuroscience of affiliation: fuss-cry. Vocalizations: parent –“motherese” high-pitched forging links between basic and clinical research on vocalizations, adult speech, none; infant – positive vocaliza- neuropeptides and social behavior. Horm. Behav. 50, 518–528. Bielsky, I.F., Hu, S.B., Young, L.J., 2005. Sexual dimorphism in the tions, fuss-cry, none. Touch: parent: affectionate, cradle, vasopressin system: lack of an altered behavioral phenotype functional, stimulatory, proprioceptive, parent stands child in female V1a receptor knockout mice. Behav. Brain. Res. 164, on his/her knees and moves infant around, touch with object, 132–136. and none; infant: intentional, accidental, none. Inter-rater Bisceglia, R., Jenkins, J.M., Wigg, K.G., O'Connor, T.G., Moran, G., reliability was computed for 15 interactions and reliability Barr, C.L., 2012. Arginine vasopressin 1a receptor gene and kappas averaged .84 (range .76–.93). maternal behavior: evidence of association and moderation. ¼ The following variables were used in the current study as Genes Brain Behav. 11, 262–268. sum proportions: Caldwell, H.K., Lee, H.J., Macbeth, A.H., Young, W.S., 2008. Vassopressin: behavioral roles of an “original” neuropeptide. Affectionate contact – was the sum proportions of parent Prog. Neurobiol. 84, 1–24. – affectionate touch and cradle. Stimulatory contact was the Carter, C.S., Pournajafi-Nazarloo, H., Kramer, K.M., Ziegler, T.E., sum proportions of stimulatory, proprioceptive, child on White-Traut, R., Bello, D., Schwertz, D., 2007. Oxytocin: knees, and touch-with-object. behavioral associations and potential as a salivary biomarker. The following “synchrony” variables were used as condi- Ann. NY Acad. Sci. 1098, 312–322. tional probabilities that index the co-occurrence of social Champagne, F.A., 2008. Epigenetic mechanisms and the behavior bin parent and infant and were computed as mean transgenerational effects of maternal care. Front. Neuroendocrinol. 29, 386–397. durations. Curley, J.P., Keverne, E.B., 2005. Genes, brains and mammalian Joint attention – parent and child are attending to the same social bonds. Trends Ecol. Evol. 20, 561–567. object. Gaze synchrony – parent and child are simultaneously Decety, J., 2010. The neurodevelopment of empathy in humans. engaged in social gaze (parent gaze at infant while infant Dev. Neurosci. 32, 257–267. gaze at parent). Douglas, A.J., 2010. Baby love? Oxytocin–dopamine interactions in The following variables were computed by lag-sequential mother–infant bonding. Endocrinology 151, 1978–1980. analysis and index the frequencies in which a specific parent Feldman, R., Gordon, I., Schneiderman, I., Weisman, O., Zagoory- behavior follows the infant's social gaze to the parent as the Sharon, O., 2010b. Natural variations in maternal and paternal care are associated with systematic changes in oxytocin following behavior that signals infant initiation of social bid (i.e., infant parent–infant contact. Psychoneuroendocrinology 35, 1133–1141. gaze to parent then parent in behavior Y). Infant gaze at Feldman, R., Gordon, I., Zagoory-Sharon, O., 2010a. The cross parent–parent increase social salience – number of times parent generation transmission of oxytocin in humans. Horm. Behav. responds by social gaze, “motherese” vocalization, or positive 58, 669–676. affect to infant social gaze. Infant gaze at parent–parent increase Feldman, R., 2012. Oxytocin and social affiliation in humans. object salience- number of times parent responds by gaze at Horm. Behav. 61, 380–391. object or joint attention to infant social gaze. Infant gaze at Feldman, R., 2007. Parent-infant synchrony and the construction of shared timing; physiological precursors, developmental parent–parent affectionate touch – number of times parent outcomes, and risk conditions. J. Child. Psychol. Psychiatry 48, responds by affectionate touch to infant social gaze. Infant 329–354. social gaze – parent stimulatory touch – number of times parents Feldman, R., Bamberger, E., Kanat-Maymon, Y., 2013. Parent- responds by stimulatory touch to infant social gaze. specific reciprocity from infancy to adolescence shapes children's social competence and dialogical skills. Attach. Hum. Dev. 5, 407–423. Feldman, R., Eidelman, A.I., 2007. Maternal postpartum behavior and the emergence of infant–mother and infant-father Acknowledgment synchrony in preterm and full-term infants: the role of neonatal vagal tone. Dev. Psychobiol. 49, 290–302. The study was supported by the German–Israeli Science Feldman, R., Eidelman, A.I., 2004. Parent–infant synchrony and Foundation (1114-101.4/2010), The I-CORE program of the the social–emotional development of triplets. Dev. Psychol. 40, planning and budgeting committee and the Israel Science 1133–1147. Foundation (#51/11), the Irving B. Harris Foundation, and the Feldman, R., Gordon, I., Zagoory-Sharon, O., 2011. Maternal and US–Israeli Bi-National Science Foundation. paternal plasma, salivary, and urinary oxytocin and parent- infant synchrony: considering stress and affiliation components of human bonding. Dev. Sci. 14, 752–761. references Feldman, R., Magori-Cohen, R., Galili, G., Singer, M., Louzoun, Y., 2011. Mother and infant coordinate heart rhythms through episodes of interaction synchrony. Infant. Behav. Dev. 34, Atzil, S., Hendler, T., Feldman, R., 2011. Specifying the 569–577. neurobiological basis of human attachment: brain, hormones, Feldman, R., Masalha, S., 2010. Parent-child and and behavior in synchronous and intrusive mothers. triadic antecedents of children's social competence: Neuropsychopharmacology 36, 2603–26015. cultural specificity, shared process. Dev. Psychol. 46, Atzil, S., Hendelr, T., Feldman, R., 2013. The brain basis of social 455–467. synchrony. Soc. Cogn. Affect. Neurosci. (in press). Feldman, R., Singer, M., Zagoory, O., 2010. Touch attenuates Atzil, S., Hendler, T., Zagoory-Sharon, O., Winetraub, Y., Feldman, infants' physiological reactivity to stress. Dev. Sci. 13, 271–278. R., 2012. Synchrony and specificity in the maternal and the Feldman, R., Weller, A., Zagoory-Sharon, O., Levine, A., 2007. paternal brain: relations to oxytocin and vasopressin. J. Am. Evidence for a neuroendocrinological foundation of human Acad. Child. Adolesc. Psychiatry 51, 798–811. affiliation: plasma oxytocin levels across pregnancy and the 132 brain research 1580 (2014) 124–132

postpartum period predict mother-infant bonding. Psychol. Olazábal, D.E., Young, L.J., 2006. Oxytocin receptors in the nucleus Sci. 18, 965–970. accumbens facilitate “spontaneous” maternal behavior in Feldman, R., 2012. Parent-infant synchrony: A bio-behavioral adult female prairie voles. Neuroscience 25, 559–568. model of mutual influences in the formation of affiliative Parke, R.D., 1996. Fatherhood. Harvard University Press, bonds. Monogr. Soc. Res. Child Dev. 77, 42–51. Cambridge, MA. Franssen, C.L., Bardi, M., Shea, E.A., Hampton, J.E., Franssen, R.A., Petronis, A., 2010. Epigenetics as a unifying principle in the Kinsley, C.H., Lambert, K.G., 2011. Fatherhood alters aetiology of complex traits and diseases. Nature 465, 721–727. behavioural and neural responsiveness in a spatial task. Ross, H.E., Freeman, S.M., Spiegel, L.L., Ren, X., Terwilliger, E.F., J Neuroendocrinol. 23, 1177–1187. Young, L.J., 2009. Variation in oxytocin receptor density in the Gamer, M., Zurowski, B., Büchel, C., 2010. Different amygdala nucleus accumbens has differential effects on affiliative subregions mediate valence related and attentional effects of behaviors inmonogamous and polygamous voles. J. Neurosci. oxytocin in humans. Proc. Natl. Acad. Sci. USA 18, 9400–9405. 29, 1312–1318. Goodson, J.L., Thompson, R.R., 2010. Nona peptide Curr. Opin. Seifritz, E., Esposito, F., Neuhoff, J.G., Lüthi, A., Mustovic, H., Neurobiol. 20, 784–794. Dammann, G., von Bardeleben, U., Radue, E.W., Cirillo, S., Gordon, I., Zagoory-Sharon, O., Leckman, J.F., Feldman, R., 2010a. Tedeschi, G., Di Salle, F., 2003. Differential sex-independent Oxytocin and the development of parenting in humans. Biol. amygdala response to infant crying and laughing in parents Psychiatry 15, 377–382. versus nonparents. Biol. Psychiatry. 54, 1367–1375. Gordon, I., Zagoory-Sharon, O., Leckman, J.F., Feldman, R., 2010b. Shahrokh, D.K., Zhang, T.Y., Diorio, J., Gratton, A., Meaney, M.J., Oxytocin, cortisol, and triadic family interactions. Physiol. 2010. Oxytocin dopamine interactions mediate variations in – Behav. 101, 679–684. maternal behavior in the rat. Endocrinology 151, 2276 2286. fl Gordon, I., Zagoory-Sharon, O., Leckman, J.F., Feldman, R., 2010c. Skuse, D.H., Gallagher, L., 2011. Genetic in uences on social – Prolactin, Oxytocin, and the development of paternal behavior cognition. Pediatric Res. 69, 85R 91R. Swain, J.E., Lorberbaum, J.P., Kose, S., Strathearn, L., 2007. Brain across the first six months of fatherhood. Horm. Behav. 58, basis of early parent-infant interactions: psychology, 513–518. physiology, and in vivo functional neuroimaging studies. Guastella, A.J., Kenyon, A.R., Alvares, G.,A., Carson, D.S., Hickie, J. Child Psychol. Psychiatry 48, 262–287. I.B., 2010. Intranasal arginine vasopressin enhances the Szyf, M., Weaver, I., Meaney, M., 2007. Maternal care, the encoding of happy and angry faces in humans. Biol. epigenome and phenotypic differences in behavior. Reprod. Psychiatry. 15, 1220–1222. Toxicol. 24, 9–19. Ishak, W.W., Kahloon, M., Fakhry, H., 2011. Oxytocin role in Thompson, R.R., George, K., Walton, J.C., Orr, S.P., Benson, J., 2006. enhancing well-being: a literature review. J. Affect. Disord. Sex-specific influences of vasopressin on human social 130, 1–9. communication. Proc. Natl. Acad. Sci. USA 103, 7889–7894. Kim, P., Leckman, J.F., Mayes, L.C., Newman, M.A., Feldman, R., Veenema, A.H., 2012. Toward understanding how early life social Swain, J.E., 2010. Perceived quality of maternal care in experiences alter oxytocin- and vasopressin-regulated childhood and structure and function of mothers' brain. Dev. socialbehaviors. Horm. Behav. 61, 304–312. Sci. 13, 662–673. Weisman, O., Schneiderman, I., Zagoory-Sharon, O., Feldman, R., Kinsley, C.H., Amory-Meyer, E., 2011. Why the maternal brain?. 2013a. Salivary vasopressin increases following intranasal – J. Neuroendocrinol. 23, 974 983. oxytocin administration. Peptides 40, 99–103. Lamb, M.E. (Ed.), 2010. The role of the father in child development Weisman, O., Zagoory-Sharon, O., Feldman, R., 2012. Oxytocin 5th ed. Wiley, New York. administration to parent enhances infant physiological and Levine., A., Zagoory-Sharon, O., Feldman, R., Weller, A., 2007. behavioral readiness for social engagement. Biol. Psychiatry Oxytocin during pregnancy and early postpartum: individual 72, 982–989. patterns and maternal–fetal attachment. Peptides 28, Weisman, O., Zagoory-Sharon, O., Schneiderman, I., Gordon, I., 1162–1169. Feldman, R., 2013b. Plasma oxytocin distributions in a large McNamara, I.M., Borella, A.W., Bialowas, L.A., Whitaker-Azmitia, cohort of women and men and their gender-specific P.M., 2008. Further studies in the developmental associations with anxiety. Psychoneuroendocrinology 38, hyperserotonemia model (DHS) of autism: social, behavioral 694–701. and peptidechanges. Brain Res. 16, 203–214. Woller, M.J., Sosa, M.E., Chiang, Y., Prudom, S.L., Keelty, P., Moore, Mosek-Eilon, V., Hirschberger, G., Kanat-Maymon, Y., Feldman, R., J.E., Ziegler, T.E., 2012. Differential hypothalamic secretion of 2013. Infant reminders alter sympathetic reactivity and reduce neurocrines in male common marmosets: parental experience couple hostility at the transition to parenthood. Dev. Psychol. effects?. J. Neuroendocrinol. 3, 413–421. 49, 1385–1395. Wynne-Edwards, K.E., 2001. Hormonal changes in mammalian Naber, F., van Ijzendoorn, M.H., Deschamps, P., van Engeland, H., fathers. Horm. Behav. 40, 139–145. Bakermans-Kranenburg, M.J., 2010. Intranasal oxytocin Zink, C.F., Kempf, L., Hakimi, S., Rainey, C.A., Stein, J.L., increases fathers' observed responsiveness during play with Meyer-Lindenberg, A., 2011. Vasopressin modulates social their children: a double-blind withinsubject experiment. recognition-related activity in the left temporoparietal Psychoendocrinilogy 35, 1583–1586. junction in humans. Transl. Psychiatry 4 (1:e3). brain research 1580 (2014) 199–218

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Oxytocin and vasopressin systems in genetic syndromes and neurodevelopmental disorders

S.M. Francisa, A. Sagarb, T. Levin-Decaninia, W. Liuc, C.S. Carterd, S. Jacoba,n aUniversity of Minnesota, Department of Psychiatry, Minneapolis, MN, USA bUniversity of California at Irvine, Department of Psychiatry and Human Behavior, USA cNorthwestern University, Feinberg School of Medicine, Chicago, IL, USA dUniversity of North Carolina, Department of Psychiatry, Chapel Hill, NC, USA article info abstract

Article history: Oxytocin (OT) and arginine vasopressin (AVP) are two small, related neuropeptide Accepted 15 January 2014 hormones found in many mammalian species, including humans. Dysregulation of these Available online 22 January 2014 neuropeptides have been associated with changes in behavior, especially social interac- Keywords: tions. We review how the OT and AVP systems have been investigated in Autism Spectrum – Oxytocin Disorder (ASD), Prader Willi Syndrome (PWS), Williams Syndrome (WS) and Fragile X Vasopressin syndrome (FXS). All of these neurodevelopmental disorders (NDD) are marked by social fi fi Autism de cits. While PWS, WS and FXS have identi ed genetic mutations, ASD stems from Prader–Willi multiple genes with complex interactions. Animal models of NDD are invaluable for Williams studying the role and relatedness of OT and AVP in the developing brain. We present data Fragile X from a FXS mouse model affecting the fragile X mental retardation 1 (Fmr1) gene, resulting in decreased OT and AVP staining cells in some brain regions. Reviewing the research about OT and AVP in these NDD suggests that altered OT pathways may be downstream from different etiological factors and perturbations in development. This has implications for ongoing studies of the therapeutic application of OT in NDD. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2014 Published by Elsevier B.V.

1. Introduction to OT and AVP neuropeptide hypothalamic origins are transported from the SON and hormones PVN to the mammalian posterior pituitary by neurosecretion where they are released into the blood stream and act as Oxytocin (OT) and arginine vasopressin (AVP) are small hormones on target tissues. In addition, both OT and AVP are mammalian neuropeptides nine amino acids in length, capable of moving throughout the central nervous system via which differ by only two amino acids. OT is produced diffusion in the cerebral spinal fluid (CSF; Neumann and primarily in hypothalamic nuclei, including the supraoptic Landgraf, 2012). The peptide-producing OT gene (OXT) is (SON) and paraventricular nuclei (PVN). AVP is also synthe- homologous with its evolutionarily related gene, vasopressin sized in the PVN and SON. In males, additional brain regions (AVP). The human OXT and AVP genes linked on chromo- including the amygdala and the bed nucleus of the stria some 20p13 are separated by only 12 kilobases of DNA, and terminalis (BNST) also produce AVP. OT and AVP of are positioned in opposite transcriptional orientations. Both

nCorrespondence to: University of Minnesota, Wallin Medical Biosciences Building, 2101 6th Street SE, Minneapolis, MN 55455, USA. Fax: 612 273 9779. þ E-mail address: [email protected] (S. Jacob).

0006-8993/$ - see front matter & 2014 Published by Elsevier B.V. http://dx.doi.org/10.1016/j.brainres.2014.01.021 200 brain research 1580 (2014) 199–218

have specific receptors, but their close evolutionary relation- example, intranasal OT (IN-OT) administration in healthy ship permits cross-talk and interacting molecular systems. human males increased prosocial behaviors and trust, espe- These neuropeptide hormones have receptors in various cially as measured by experimental economic games brain regions and throughout the body, including areas that (Baumgartner et al., 2008; Kirsch, 2005; Kosfeld et al., 2005). are important for regulating social behavior and reactivity to IN-OT may also increase gaze towards the eye region of the stressors. face (Guastella et al., 2008), and has been associated with In both, the human and mouse genomes OT and AVP improved facial memory (Rimmele et al., 2009), enhanced neuropeptide genes are located adjacently on the same chro- salience of social cues (Shamay-Tsoory et al., 2009), and mosome. Often the blood levels of both hormones are highly improved performance on the reading the mind in the eyes correlated (Dai et al., 2012), suggesting a coordinated release. (RMET) task (Domes et al., 2007). The receptors for both neuropeptides are localized in specific As previously reviewed, OT has been found to have areas of the nervous system, especially in the brainstem. anxiolytic effects improving social interactions, reducing fear, These brain regions influence social and adaptive behaviors, and improving the ability of healthy volunteers to interpret as well as regulate the hypothalamic–pituitary–adrenal axis subtle social cues (Macdonald and Macdonald, 2010). In (HPA) and autonomic nervous systems (Lim et al., 2005, 2004). addition, OT dysfunction has been associated with neurop- Because OT and AVP are closely related and have the ability to sychiatric disorders such as autism in human studies (Domes act on the other's receptors, it has been proposed that they et al., 2007; Ishak et al., 2011; Winslow and Insel, 2004). By evolved to interact and sometimes have opposing physiologi- 2010 there were over 20 OT administration studies, which cal effects. For example, both hormones have been shown to included ASD, schizophrenia, postpartum depression, post- affect the control of the autonomic nervous system, with OT traumatic stress disorder (PTSD), and irritable bowel syn- having primarily parasympathetic actions and AVP serving as drome (Macdonald and Macdonald, 2010). There have been a a central and peripheral regulatory component of the sympa- growing number of studies investigating the ability of IN-OT thetic nervous system and HPA axis (Kenkel et al., 2012; to treat a range of neurobehavioral disorders due to the Sawchenko and Swanson, 1985). However, at high levels the associations between IN-OT and alterations in social deci- neuropeptides can be partial agonists for their homologous sion-making, processing of social stimuli, certain social receptors, which may result in AVP and OT pathways inter- behaviors such as eye contact, and social memory. acting (Chini et al., 1996). Of particular importance in neurodevelopmental disorders (NDD) is the fact that OT and AVP can modulate social and repetitive behavior and other manifestations of anxiety and 2. Autism spectrum disorders state regulation (Carter, 2007). Animal research has generally associated OT release or exposure with positive sociality, In 1943, Leo Kanner described a male patient with repeti- reduced anxiety, and lower levels of reactivity to stressors tive behaviors—“stereotyped movements [and]…repeti- (Carter, 1998; Neumann and Landgraf, 2012). AVP influences tions carried out in exactly the same way in which they anxiety, the regulation of HPA and stress responses. In had been performed originally” and difficulties with social general, central AVP is described as anxiogenic (Landgraf communication—“he always seemed to be parroting what and Wigger, 2003). However, there is also evidence in rats he had heard said to him at one time or another…Words to that the effects of AVP are brain region specific and dose- him had a specifically literal, inflexible meaning. He dependent. For example, AVP may be anxiolytic if given in seemed unable to generalize, to transfer an expression to low doses (Appenrodt et al., 1998). another similar object or situation” (Kanner, 1943). This Mouse knockout (KO) studies of the OT receptor (OXTR) or group of symptoms, later extended and described in detail, OT regulators have found decreased social memory or recog- is currently known as ASD. As described in the DSM-5 nition (Ferguson et al., 2000; Jin et al., 2007; Takayanagi et al., (American Psychiatric Association, 2013), ASD is character- 2005). Oxtr KO mice also displayed decreased cognitive flex- ized by persistent deficits in social communication and ibility and a resistance to change of a learned pattern of social interaction across multiple contexts, and the diag- behavior that is comparable to restricted/repetitive interests nosis requires the presence of restricted, repetitive patterns (Sala et al., 2011). Both social deficits and behavioral rigidity of behaviors, interests, or activities. ASD is a heritable were ameliorated by OT administration (Sala et al., 2011). The (Bailey et al., 1995)andhighlyheterogeneousdisorder, finding that OT continues to have effects in Oxtr KO mice caused by familial genetic risks in addition to possible supports the hypothesis that OT can influence behavior gene-environment interactions during early development through other receptors, especially the AVP receptors (Chaste and Leboyer, 2012). Individuals with ASD often (e.g. AVPR1A and/or AVPR1B). Given the influence of these suffer with anxiety disorders, irritability or aggression, neuropeptides on brain regions affecting both social and and come to clinical attention due to their difficulties at repetitive behaviors, modulation of OT and AVP pathways home and school related to their communication deficits are being explored as treatment targets for disorders, includ- and restricted interests. Unfortunately there are currently ing Fragile X syndrome (FXS) and Autism Spectrum no approved medications to treat the social deficits or Disorders (ASD). restricted, repetitive behaviors (RRB) that are the core This and other research has set the stage for a series of symptoms of ASD. There is some evidence in animal and recent studies on the effects of exogenous OT treatments in human studies that OT improves the core symptoms humans (Ebstein et al., 2012; Macdonald and Feifel, 2013). For of ASD. brain research 1580 (2014) 199–218 201

2.1. Intranasal and intravenous OT studies in ASD tossed the ball back to the individual) that was similar to the control subjects also performing the task. This preference Currently medications for ASD concentrate on alleviating was further supported by the patients' reporting of trust, certain symptoms, but not the core features of ASD. Risper- towards the “good player” after OT administration. Andari idone and aripiprazole may be used for irritability, whereas and colleagues also found that after IN-OT, patients increased guanfacine and clonidine are used off label for aggression, and eye gazing time on the socially informative region of the face. selective serotonin reuptake inhibitors (SSRI; i.e. escitalopram, While single dose studies are valuable, evaluating long term fluoxetine, and sertraline) are used to treat anxiety (Jaselskis effects of OT are essential to determine if OT has a ther- et al., 1992; McCracken et al., 2002; Owley et al., 2010). Recently, apeutic potential in ASD (Macdonald and Feifel, 2013). OT has been investigated as a target the treatment for ASD Recently a five-day OT administration study was conducted core symptoms, social deficits and RRB. Defined by DSM-5, during parent–child interaction training (Dadds et al., in press). restricted, repetitive patterns of behavior include stereotyped Individuals with high functioning ASD received 12 or 24 IU or repetitive motor movements, insistence on sameness, (depending on the weight of the patient) IN placebo or IN-OT. inflexible adherence to routines, or ritualized patterns of The OT or placebo was administered once daily and RRB, verbal or nonverbal behavior. Highly restricted, fixated inter- emotion recognition, social interaction skills, and general ests that are abnormal in intensity or focus, and hyper- or behavioral adjustment were assessed. While improvements hyporeactivity to sensory input or unusual interest in sensory over time were detected in both OT and placebo, there were no aspects of the environment are also RRB. differences observed between the two groups. Several pro- Several studies, using intravenous OT or IN-OT, in patients posed possible explanations for these null findings were: with ASD have been conducted (Table 1). It has been shown (1) emotion recognition was measured pre-post changes fol- that nonapeptides, like AVP, can be measured in CSF after lowing multiple exposures versus while the patient was under intranasal administration (Born et al., 2002). Ease of giving the influence; (2) lower-order RRB respond to OT (Hollander intranasal drugs makes it preferred for most ASD studies, et al., 2003), while higher-order RRB do not (Anagnostou et al., although more research needs to be conducted on how IN-OT 2012); (3) increased eye gaze frequency is usually measured reaches the brain and how it regulates receptors and neural with artificial or computerized faces, while they had “real-life” pathways with different or chronic dosing strategies. Several interactions; (4) the OT receptor system disruptions in some studies have measured OT responses to single dose challenges patients with ASD may respond differently than in other ASD in ASD (Andari et al., 2010; Guastella et al., 2010), while few patients; and (5) differences between the studies regarding age have examined longer term treatment effects (Anagnostou and diagnostic characteristics of the sample. Studies pairing et al., 2012). With varying administration and duration study OT with a therapeutic activity or social training are highly protocols, studies have often focused on symptom subdo- needed, although design and outcome measures across stu- mains or defined social tasks including: RRB (Hollander et al., dies will need to be similar in order to better interpret and 2003), emotion recognition (Dadds et al., in press; Guastella compare results in ASD participants. et al., 2010), affective speech comprehension (Hollander et al., Investigating adults with ASD, Anagnostou et al. (2012) 2007), and facial recognition (Domes et al., 2013). studied the safety and therapeutic effects of IN-OT with Single dose studies, or challenges, have been utilized to respect to two core symptom domains: social cognition and study the acute and immediate effects of OT. An initial study functioning, and RRB. They performed a randomized, double- in ASD examined the effects of a four-hour continuous dose blind, placebo-controlled parallel trial of IN-OT versus pla- of intravenous OT (Hollander et al., 2003). After one hour of cebo. This was the first study to employ a treatment trial of infusion there was a decrease in RRB (repeating and touch- daily administration of IN-OT in ASD. Overall, the IN-OT was ing). After four hours, 13 patients (86.7%) versus six control well tolerated when given daily and no serious adverse subjects (40%) had decreases in RRB. This study demon- effects were reported. This pilot study suggested therapeutic strated that administration of OT led to a decrease in a core potential with daily administration of IN-OT in this popula- ASD symptom, RRB. More recently, a double-blind, rando- tion of adults with ASD. This six-week study noted improve- mized, placebo controlled study of IN-OT in 16 males with ments in social cognition, quality of life, RRB, and some ASD (ages 12–19 years old) showed that a single IN-OT dose measures of emotional well-being with IN-OT, in essence, could improve the ability to recognize emotion, particularly improvement in the core domains of ASD. It may be impor- in easy queries (Guastella et al., 2010). It is unclear whether tant to note that reports of individual differences in the emotion recognition performance goes back to baseline or response to IN-OT are increasingly observed, although sam- pre-OT exposure after a single dose or if there are long-term ple sizes in studies will need to be larger to explore individual learning effects. response variation. Andari et al. (2010) also performed a single dose IN-OT study in 13 individuals with ASD and in age-matched controls 2.2. OT/AVP plasma levels in ASD to study the effect on social deficits. They researched the effects of IN-OT on trust and preference using a social ball The inability to directly access the brain's oxytocinergic tossing game (for greater detail of this task please see pathways has constrained human research. Therefore, per- supplementary section of Andari et al., 2010). This group also ipheral OT levels have been used as proxies for OT levels in assessed the visual scanning of faces under the influence of the brain. A widely used measurement is plasma OT, IN-OT. They found that patients given IN-OT had a significant although urine and salivary OT levels have also been preference for the “good player” (the computer player who explored in some studies. A range of studies have observed 202 brain research 1580 (2014) 199–218

Table 1 – A summary of OT human trials in ASD.

Article Design Sample Dose Assessments, Outcome measurments and/or tasks

Hollander Randomized 15 Adults with Initial vial of Repetitive Behaviors: Observed a et al. (2003) double-blind, Asperger Syndrome 10 μ/mL Evaluated using an significant reduction placebo-controlled or Autism (14 males) combined with a in-lab method that in repetitive crossover design Mean Age 32.9 1L bag of normal assessed six behaviors over time ¼ years (range saline given at a restricted, repetitive versus placebo 19.4–55.6 years) scheduled behaviors (RRB) at session; 86.7% of the increased rate of baseline, 60, 120, 180, participants during infusion (4 and 240 minutes the OT session had continuous reduction in RRB, hours) while 40% of the subjects had a reduction in RRB during the placebo session (non- significant)

Andari et al. Randomized 13 Adults with 24 IU IN-OT Cyberball Game Adults with Asperger (2010) double-blind, Asperger Syndrome (social interaction Syndrome or High placebo-controlled or High Functioning task) Functioning Autism within-subjects Autism (11 males) exhibited stronger design interactions in a simulated ball game (similar to healthy adults), and gazed at eyes longer on a face task with IN-OT Mean Age 26 years Free viewing of ¼ (range 17–39 years) pictures of faces

13 Age-matched, healthy adults tested in a single visit (11 men)

Mean Age 26 years ¼ (range 18–40 years)

Guastella Randomized 16 Young males with 18 IU (ages 12–15 Reading the Mind in Significant et al. (2010) double-blind, Asperger Syndrome years) or 24 IU the Eyes Task (RMET) improvement in placebo-controlled or Autism (ages 16–19 years) RMET compared to crossover design IN-OT placebo Significant improvement in identifying emotions of easier items on the test Mean Age 14.88 ¼ years (SD 2.42 ¼ years)

Anagnostou Randomized double- 19 Adults with ASD 24 IU IN-OT Social Function/ Primary Outcome et al. (2012) blind, placebo- (16 males) Cognition: Clinical Measures: A non- controlled, parallel Global Impression, significant decrease design trial of IN-OT DANVA (primary in lower-order RRB; vs. placebo measures); RMET, 30% of the IN-OT Social group rated as Responsiveness Scale improved by CGI, (secondary measures) while 11% of the placebo participants improved (non- significant) Mean Age 33.20 Repetitive Behaviors: Secondary Outcome ¼ years (SD 13.3 RBS-R (primary Measures: A ¼ years) measure); Yale significant Brown Obsessive improvement in brain research 1580 (2014) 199–218 203

Table 1 (continued )

Article Design Sample Dose Assessments, Outcome measurments and/or tasks

Compulsive Scale social cognition as (secondary measure) measured by RMET and in quality of life as measured by WHOQOL

Quality of Life: WHO Quality of Life Questionnaire (WHOQOL; secondary measure)

Dadds et al. Randomized double- 38 Young males with 12 IU (o40 kg) Social Interaction: No improvement in (in press) blind, control trial ASD and a high level or 24 IU (>40 kg) Social Skills rating social interaction, over five days of commonly IN-OT Scale, Video micro- emotion recognition, comorbid disorders coding and global or general behavioral coding of adjustment observations

During the five days, OT Mean Age 11.79 Repetitive Behaviors: ¼ the participants and years (SD 2.82 Social Reciprocity ¼ their families years) Scale, Micro-analysis underwent parent- of observation videos child interaction training, which Placebo Mean Emotion Recognition: included Age 10.74 years UNSW Facial ¼ Mindreading (SD 2.38 years) Emotion Task ¼ program and video feedback (based on Generalized Effects/ Video Interactive Diagnostic Status: Guidance) OSU, Childhood Autism Ratings Scale, DISCAP-ASD

associations between peripheral OT/AVP levels and social the extended form (OT-X) is cleaved by enzymatic activity to stimuli (Kenkel et al., 2012; Schneiderman et al., 2012; yield the active peptide OT. There was also a positive Schradin et al., 2013; Seltzer et al., 2010; Wismer Fries et al., correlation between OT-X and checklist items associated with 2005; Zhong et al., 2012). ASD including stereotypies; OT-X correlated negatively with As early as 1996, it was suggested by a number of an item describing abnormalities in comfort giving within the researchers, including Waterhouse et al. that dysfunction in ASD group. Consequently, changes in OT processing, specifi- the OT and AVP systems might contribute to the atypical cally a failure to completely process the prohormone OT-X, social behaviors in ASD. Two years later by studying the OT might lead to a deficiency in OT, thus exacerbating some of plasma levels from 29 ASD and 30 age-matched typical the symptoms of ASD, such as features of social deficits. To control children, Modahl et al. (1998) reported low levels of our knowledge this study has not been replicated. Further- plasma OT in the children with ASD. In another study, which more, other studies often in older patients have failed to utilized Wing's topology, children were classified as “aloof”, report an OT deficiency (Jansen et al., 2006; Miller et al., 2013). “active but odd” and “overly formal” (Leekam et al., 1997). The Future studies may benefit from the measurement of OT-X in lowest OT levels were found in the "aloof" subgroup. This addition to plasma OT given the diverse methods employed suggested that the most severe socially-aloof symptoms were for assaying OT, and the finding that the failure to process associated with more OT dysfunction. this prohormone can lead to a deficiency of OT. Building on these results with the same study sample, Recently, connections between peripheral OT/AVP levels Green et al. (2001) conducted another OT study and examined and ASD were investigated (Miller et al., 2013). Miller et al. the forms of the OT peptide found in affected and unaffected measured OT and AVP plasma levels in 75 boys and girls children. Their results showed that there was an increase in (40 high-functioning ASD, 35 typically developing) aged OT-X, the precursor for OT, as well as an increase in the ratio 8–18 years. Miller et al. not only reported associations of OT-X/OT associated with the reduction in OT observed in between the plasma levels and ASD behaviors, but sex the patients with ASD. During the normal production of OT, differences as well. Higher levels of OT were observed in all 204 brain research 1580 (2014) 199–218

girls, and all boys had significantly higher levels of AVP. The et al. (1998) performed plasma extractions and then radio- higher OT levels were associated with greater anxiety in all immunoassays (RIA) whereas Miller et al. (2013) utilized an girls, and with better pragmatic language in all subjects. enzyme immunoassay (EIA) with different plasma prepara- Gender differences were also noted within the ASD sample. tion methods. Additionally, there is also specific lab gener- A positive association between AVP levels and RRB was ated RIA versus commercial EIA and RIA kits. When reported in ASD girls, although a non-significant association manufacturer instructions are followed, values obtained have with RRB was found in boys with ASD. Because of the limited a similar order of magnitude, but it has been noted that some number of girls affected by ASD, few other studies have of these kits may also be detecting closely related metabo- sampled a large enough sample size of girls to investigate lites. Note that studies often prepared samples differently gender differences in OT plasma levels or response to with varying plasma processing/extraction methods and use exogenous administration. of different assay techniques. Future research will need to Some studies have examined OT levels in addition to determine if differences observed in the resultant OT levels of other hormones and blood biomarkers. In a study of adults ASD studies reflect differences in the study populations (i.e. with ASD, basal OT levels and heart rate were elevated in the age) and/or the methods for assaying OT. ASD is a very ASD group compared to healthy controls. These adults with heterogeneous disorder and OT level differences may be ASD showed normal cortisol responses to a public speaking specific to clinical and etiological subgroups within the task, but no change in norepinephrine, epinephrine, OT or broader ASD population. In addition, there is variability of AVP (Jansen et al., 2006). Recently, Hammock et al. (2012) OT plasma levels across typical and healthy populations. The analyzed correlations between the biomarkers of plasma OT inherent U-shaped distribution (Zhong et al., 2012) observed and whole-blood serotonin (5-HT) levels in children and in normative populations may also add to variability seen in adolescents diagnosed with ASD and not on medications. OT measurements in ASD studies. Animal studies have shown that OT and 5-HT influence each other's release (Bagdy and Kalogeras, 1993; Jorgensen et al., 2.3. OT and AVP animal and genetic studies in ASD 2003; Yoshida et al., 2009) and there have been many reports of hyperserotonemia within a subgroup of individuals with The search for genes and biological risk factors contributing ASD (Abramson et al., 1989; Chugani et al., 1999; Kuperman to ASD and its core symptoms has resulted in a range of et al., 1985; Leboyer et al., 1999; Leventhal et al., 1990; Schain human and animal model studies. Recently, several research- and Freedman, 1961). OT and 5-HT were negatively correlated ers have examined how the OT system is altered in various with each other in the Hammock et al. (2012) study. Whole animal model or influences social and repetitive behaviors. blood 5-HT was found to be negatively correlated with age, For example, the BTBR T tf/J (BTBR) mice have low social þ having lower levels in adolescence than in childhood. This interactions, decreased vocalization in social settings and OT/5-HT relationship was especially prominent in children increased levels of repetitive self-grooming, behavioral phe- younger than 11 years old. Age may be an important covari- notypes similar to the core symptoms of ASD. Comparing ate, because some studies do not find OT deficiencies or have BTBR to the standard inbred highly sociable mouse, C57BL/6J reported higher than expected OT levels in their ASD samples (B6), Silverman et al. (2010) found elevated OT in the PVN, (Jansen et al., 2006; Miller et al., 2013). OT like 5-HT may plasma corticosterone (in the trunk), and change after puberty according to recent data (Hammock receptor (GR) mRNA in CA of the hippocampus in the BTBR. et al., 2012). Measurements of the other neurochemicals (CRF in PVN, GR Interest in parental bonding has led to the study of OT mRNA in CA2 and PVN) showed no differences between the levels in parents of typical children, and more recently strains. parents of children with ASD. An association between per- Another model, the BALB/cJ, has low sociability across ipheral OT and parental care, both maternal and paternal, development (Brodkin, 2007). A substrain of the BALB/cJ, the was reported by Feldman et al. (2012). When comparing BALB/cByJ mouse, was also characterized as a good ASD parents and non-parents, parents were found to have higher model (Brodkin, 2007; Moy et al., 2007). These mice displayed levels of OT. Higher plasma OT levels were also associated low sociability with intact olfaction, locomotor activity and with longer durations of gaze synchrony and reporting of relatively high levels of anxiety. Similarly, the C58/J mouse is greater parental care during the parent's childhood, while another ASD mouse model described in 2010 by Ryan et al. lower plasma OT corresponded to less parental touch. Sub- (2010). These mice show low sociability and deficits in social sequently, Xu et al. (2013) published a study comparing the communication. Most striking, however, is their abnormal OT and AVP plasma levels of mothers with and without ASD RRB including increased rates of pivoting, back flipping, children in a Han population. They found that the mothers of upright scrabbling and “jack-hammer” jumping. All the traits ASD children had significantly lower plasma OT/AVP com- are found in both the males and females of the C58/J strain. pared to the control mothers, as well as, a significant Recently, the reactions of the BALB/cByJ and C58/J strains to correlation between the plasma levels of the neuropeptides OT administration were compared (Teng et al., 2013). and the child's autistic behavior scores. Teng et al. (2013) did not only note the different reactions Over the last few years studies measuring peripheral OT to OT administration between BALB/cByJ and C58/J, they also have increased. As discussed in McCullough et al. (2013) and studied whether acute versus subchronic administration had Szeto et al. (2011), the methodologies can lead to vastly differing outcomes (see paper for drug and experimental different results (increased values, decreased values or values timeline). Both strains were administered OT peripherally differing by an order of magnitude). For example, Modahl via intraperitoneal administration. Acute administration in brain research 1580 (2014) 199–218 205

BALB/cByJ showed no change in sociability, and RRB were not group given the results of previous studies (Jemel et al., 2006; assessed in this timeline. However, in the subchronic regi- Schultz, 2005). Unexpectedly, they saw increased activation men BALB/cByJ mice displayed a significant increase in social in the fusiform brain regions. This led them to conclude that behaviors. Subchronic OT administration in the C58/J model while more research needs to be done to confirm their also induced prosocial effects. In male mice these effects findings, the link between CD38, the processing of social appeared two weeks post treatment, but the prosocial effects information and ASD was further solidified. were evident in female mice sooner. Regarding RRB, Teng Other researchers have looked directly at the OXTR gene, to et al. observed a decrease in repetitive behaviors with find possible genetic links between OT and ASD. Wu et al. (2005) increased self-grooming after an acute single dose of OT, in genotyped four SNPs across OXTR in 195 Chinese Han ASD trios. the C58/J mice. Most notably these studies have provided With Family Based Association Testing (FBAT), they revealed insight into how the genetic heterogeneity observed in significant associations between ASD and two OXTR SNPs humans may account for the wide variety and degree of (rs2254298, rs53576). When the markers were combined to behaviors seen in ASD. Additionally, the Teng et al. study create haplotypes, significant associations were found for all highlighted how treatments can be dependent on both geno- markers and especially in haplotypes containing rs53576. Fol- type and dose regimen. It is striking that all three mouse lowing up the Wu et al. study, Jacob et al. (2007) researched models of ASD have alterations in the OT system or respond to rs53576 and rs2254298 in a Caucasian sample with strictly OT administration. This suggests that OT may be affected defined autism. They genotyped the OXTR SNPs in 57 autism downstream in strains of mice that have different etiological trios. In this sample a significant association was observed factors influencing social and repetitive behaviors. between diagnosis and rs2254298. While the G allele was more As the animal models have shown, the genetic hetero- frequent than the A allele, of note was the overtransmission of geneity of individuals with ASD could also account for the the G allele to the autistic Caucasian probands versus over- complexity of the disorder's genetic etiology. In a recent transmission of the A allele in the Chinese Han sample. review (Ebstein et al., 2009), genetic polymorphisms of recep- Additional studies by Yrigollen et al. (2008) and Campbell tor and pathway regulators of OT and AVP, such as AVPR1a, et al. (2011) continued to build an association between the OT OXTR, neurophysin I and II, and CD38 were discussed. In this system and ASD. Yrigollen and colleagues hypothesized that review, Ebstein and colleagues presented preliminary data genes associated with affiliative, social, and/or bonding beha- regarding their findings about CD38, a transmembrane gly- viors would also be associated with ASD and several of its coprotein involved in OT secretion and associated with OT symptoms. Therefore, they studied 177 ASD probands from 151 plasma levels. The group genotyped 12 tag single nucleotide families and they found the statistical strength was in the OXTR polymorphisms (SNPs) across CD38 in 170 ASD trios. IQ and results. Different OXTR SNPs were significantly associated with social skills via Vineland Adaptive Behavior Scales (VABS) stereotyped behaviors, communication skills, the multivariate were assessed in the sample. They found a significant ADI phenotype and multi-measurement variable overall diag- association between CD38 SNPs and categorical ASD mea- nosis. Also, there was a significant SNP in the OXT/AVP region sures, assessed by the Autism Diagnostic Interview – Revised associated with stereotyped behaviors. In 2011, Campbell and (ADI-R; Lord et al., 1994) and Autism Diagnostic Observation colleagues, utilizing a repository sample of 2333 individuals Schedule – Generic (ADOS-G; Lord et al., 2000). Significance with ASD in 1238 pedigrees, analyzed 25 markers across OXTR was also observed between VABS scores and four CD38 SNPs, and detected associations. Although these SNPs had previously haplotypes and VABS, and CD38 mRNA levels and VABS. Two been associated with ASD diagnosis in studies, they also looked studies in 2010 further supported a role for CD38 in ASD. Lerer at associations with subphenotype measurements of social and colleagues (2010) noted a reduced expression of CD38 in impairments in ASD and found associations with their selected the lymphoblastoid cells of patients with ASD as compared to three SNPs and ADOS, ADI-R and SRS measurement. While the “unaffected” parents. Then Munesue et al. (2010), observed a results of the studies indicate a need for further research, these CD38 SNP in association with high functioning ASD in some genetic studies have laid the foundation linking OT, social populations. In 2012, several CD38 studies were completed tasks, and more broadly ASD. that looked at the association of CD38 and ASD. In particular, Sauer et al. (2012) researched the common CD38 variant, rs3796863, in healthy young men. This SNP stood out in the 3. Prader–Willi syndrome and OT ASD association studies performed by Lerer et al. (2010) and Munesue et al. (2010). Data were attained in a double-blind Prader–Willi syndrome (PWS) is a complex disorder with placebo-controlled crossover design using IN-OT. The sub- multisystem effects and a distinct behavioral phenotype. It jects performed two tasks following administration of OT or occurs in approximately 1/10,000–1/30,000 births, and is placebo: (1) a face matching task, and (2) a gaze processing initially characterized by severe infantile hypotonia and task. They found that the men with the ASD risk allele had difficulty feeding, although later in infancy and into adoles- significantly slower reaction times (RT) during the face cence individuals with PWS often eat excessively and develop matching task and that it was specific to social versus non- morbid obesity. Other characteristics of PWS include hypo- social stimuli. IN-OT reduced RT in the risk group. Gaze gonadism, short stature, small hands and feet and strabis- processing did not yield any significant results. Functional mus. The cognitive phenotype is marked by delayed motor MRI data also were attained as the subjects performed the and language development, and behavioral difficulties tasks. Sauer et al. (2012) hypothesized decreased activity in including compulsive behavior, stubbornness and temper the amygdala and fusiform brain regions of the risk allele tantrums (Bittel et al., 2007b; Cassidy et al., 2011). The many 206 brain research 1580 (2014) 199–218

behavioral and psychiatric manifestations of PWS are evident on infants with hypercalcemia, developmental delays, car- in early childhood, and are characterized by hyperactivity, diac malformations and dysmorphic facial features (Morris, impulsivity, temper tantrums, emotional lability, anxiety and 1993). However, better characterization of this syndrome has repetitive behavior (Borghgraef et al., 1990; Gross-Tsur et al., elucidated a distinct behavioral phenotype marked by an 2001; Whitman and Accardo, 1987). Often this phenotype is increased social drive paired with social fearlessness, poor suggestive of ASD as well as attention deficit hyperactivity judgment, difficulty forming peer relationships and high disorder (ADHD; Cassidy et al., 2012). Face processing is also anxiety levels (Jarvinen et al., 2013). The cause of this disorder altered in individuals with PWS, as they have difficulty has been determined to be a deletion of 25–30 genes in the reading facial expressions (Whittington and Holland, 2011). q11.23 region of either maternal or paternal chromosome 7 The cause of PWS is the lack of expression of specific that spans approximately 1.5 megabases (Ewart et al., 1993; paternal genes located on chromosome 15q11.2-q13. Many of Korenberg et al., 2000; Lowery et al., 1995; Schubert, 2009). the genes expressed in this region come from the father, as ELN, the gene for elastin, was the first deleted gene identified those from the mother are normally inactivated. Conse- and its absence is indicative of a diagnosis of WS. While ELN quently, either a lack of expression or absence of the paternal disruption affects connective tissue, particularly of the aorta copy of the genes in this region leads to no expression (Saitoh (Lowery et al., 1995), other genes such as LIMK1, CYLN2, GTF2I et al., 1997). This may occur through microdeletions in the and GTF2IRD1 are involved in the behavioral phenotype of WS paternal chromosome, no copy of the paternal chromosome (Jarvinen-Pasley et al., 2008). The deletion of Gtf2i as well as paired with two copies of the maternal chromosome—uni- Gtf2ird1 has been shown to be involved in the social pheno- parental disomy (UPD), or imprinting defects due to epige- type specifically (Proulx et al., 2010; Sakurai et al., 2011). netic causes (Cassidy et al., 2012). The genes expressed in this The social phenotype associated with WS is striking due to region have been studied at length to develop models of PWS the hypersociability of the affected individuals, as well as the and to delineate their roles in the different aspects of the preference for novel social over non-social stimuli (Jarvinen- PWS phenotype. Such studies are complicated by differences Pasley et al., 2008; Jarvinen-Pasley et al., 2010) and increased in the behavioral phenotype between individuals with dele- eye contact (Mervis et al., 2003). In addition, the speech of tions and those with UPD, as those with UPD have a less individuals with WS is marked by high levels of socially severe phenotype (Bittel et al., 2007a) and higher verbal IQ engaging language as compared to controls or individuals scores (Dimitropoulos et al., 2000). with other developmental disorders such as Down Syndrome While the deletion of no one individual gene has been found (Jarvinen-Pasley et al., 2010; Jarvinen et al., 2013). However, to cause PWS, research has shown that the lack of expression of this does not translate into the development of social rela- multiple genes may be central to the syndrome's expression. tionships as individuals show difficulty with social adjust- Specifically, five polypeptide coding genes, namely MKRN3, ment (Gosch and Pankau, 1994, 1997) and social judgment MAGEL2, MAGED1, NECDIN and SNURF-SNRPRN,havebeen (Einfeld et al., 1997; Gosch and Pankau, 1997). In addition, shown to be centrally involved in PWS. Animal models lacking affected individuals show deficits in social understanding, as one of these genes have been developed for Magel2 (Boccaccio evidenced by difficulty identifying affect (Gagliardi et al., et al., 1999), Maged1 (Dombret et al., 2012), Necdin (Lavi-Itzkovitz 2003; Plesa-Skwerer et al., 2006) or other's mental states et al., 2012; Muscatelli et al., 2000)andSnurf (Tsai et al., 1999), (Jarvinen-Pasley et al., 2008). although none of these individual gene disruption models The high sociability of individuals with WS positions this completely recapitulates the PWS phenotype. syndrome as a good mechanism through which to under- Another line of approach to elucidate the physiological stand the biological underpinnings of social behavior. Mouse underpinnings of PWS has been to examine the OT system models of WS include GTF2I deficient mice which display in individuals with PWS as well as in animal models. There is a increased social interaction with novel mice and diminished deficit of OT producing neurons in the PVN of persons with PWS social habituation (Sakurai et al., 2011) and Gtf2ird1 deletions, (Swaab et al., 1995), as well as lower levels of OT in CSF (Martin which also show increased sociability (Proulx et al., 2010). et al., 1998). IN-OT administration increases trust in others and Recently, de novo duplications of regions of 7q11.23 have decreases disruptive behavior in individuals with PWS (Tauber been shown to be associated with ASD, whereas deletions of et al., 2011). In addition, administration of OT has also been the same region lead to WS (Sanders et al., 2011). Such shown to rescue behavior in a Maged1 deletion model of PWS in opposite effects of gene expression leading to markedly which there is a decrease in hypothalamic OT (Dombret et al., contrasting phenotypes raises the issue of dosage effects, 2012). Although rescue was not attempted in the Necdin model, but it should be noted that both ASD and WS phenotypes this mutant also shows a reduction in OT-producing neurons in include abnormal social relationships, although through the hypothalamus (Muscatelli et al., 2000). Consequently, there different mechanisms. Whereas individuals with WS show appears to be disruption of the OT system in individuals with prolonged face gaze, those with ASD display reduced face PWS, which is recapitulated in different animal models. How- gaze (Riby and Hancock, 2009). In addition, although children ever, the exact mechanism of OT dysregulation is unclear. with WS and ASD display high levels of anxiety, individuals with ASD have higher levels of RRB as well as greater rates of social phobia and separation anxiety (Cascio et al., 2012). 4. Williams syndrome and OT As deletions or increased expression of genes in the region defining WS can lead to the contrasting phenotypes of WS or Williams syndrome (WS) was first described over 50 years ago ASD, respectively, the possibility of dysregulation of OT was (Williams et al., 1961). The first reported cases were focused examined by Dai et al. (2012). They show increased baseline brain research 1580 (2014) 199–218 207

levels of OT in individuals with WS as compared to controls. spines of neocortical pyramidal neurons of FXS subjects are Additionally, OT levels correlated positively with increased longer and thinner than those of matched controls, indicating approach to strangers as well as decreased adaptive social perhaps the spines fail to mature normally in FXS patients behaviors. These results suggest that there may be a dose (Hinton et al., 1991; Irwin et al., 2001; Rudelli et al., 1985; dependent effect of OT, as high levels may impair adaptive Wisniewski et al., 1991). social behavior and may partly underlie the maladaptive The majority of FXS patients have social anxiety and social phenotype of WS. almost a third have symptoms that overlap with ASD (Hagerman et al., 2010). Published studies have reported the prevalence rate of FXS and autistic behaviors/ASD diagnosis 5. Fragile X syndrome to range from 25–47%, however sample sizes are often small (Hatton et al., 2006; Morton et al., 1997). Like FXS and ASD, Named for the fragile site observed at Xq27.3, Fragile autistic symptoms are more common in males than females. X Syndrome (FXS) is the most common inherited form of Individuals with both FXS and ASD often have poorer devel- intellectual disability and the most common known single opmental outcomes, lower cognitive abilities, lower levels of gene mutation associated with ASD (O'Donnell and Warren, adaptive behavior and more problem behaviors than FXS 2002). Prevalence estimates range from 1 case in 1000 to 1 individuals with fewer autistic behaviors. Of the individuals  case in 4000 males and has settled at 1 case in 6000 world- with FXS and autistic behaviors, 15–40% of males and a few wide for females (Brown, 1990; Morton et al., 1997; Turner females meet the diagnostic criteria for ASD (Berry-Kravis et al., 1996; Webb, 2010). This rare genetic disorder is char- et al., 2002). This group also tends to present with more acterized by specific physical features, as well as cognitive severe communication deficits, stereotyped behaviors, and and behavioral phenotypes (Berry-Kravis et al., 2002, 2011; social anxiety versus social disinterest. Additionally, males McLennan et al., 2011). The physical features can include: a present with more severe developmental delays than long narrow face with large protruding ears, connective tissue females. Overlapping behaviors between ASD and FXS, such abnormalities (i.e. hyperextensive ), macroorchidism, as eye gaze avoidance (Hall et al., 2009), have led many macrocephaly, obesity (especially in young males), loose skin scientists to study FXS as a way to understand and possibly over the hands, a high arched palate, a vertical plantar crease target treatment for ASD. Some of the medical problems and flat feet (Moy et al., 2009; Schapiro et al., 1995). The exhibited within this population include seizures (15–20% of behavioral and social characteristics of FXS include: hyper- male children, usually limited to childhood), gastroesopha- activity, attention difficulties, mood lability, compulsive and geal reflux, failure to thrive in early infancy, hypotonia, perseverative behaviors, some aggressive outbursts, learning recurrent otitis media and sinusitis, vision problems, cardiac deficits, developmental delays (including delayed speech valve prolapse, sleep disorders, and orthopedic issues and development), social shyness and gaze avoidance, sensory dental malocclusions. In the few girls studied with FXS, these hypersensitivity and withdrawal from touch, stereotypic medical problems are more variable (Berry-Kravis et al., 2002). movements and behaviors (i.e. hand flapping and rocking), Diagnosis is based on DNA analysis that identifies the poor motor coordination and echolalia (Hagerman et al., 2009; number of CGG repeats in the fragile X mental retardation 1 Hall, 2009; Hall et al., 2009; Moy et al., 2009). Many of these (FMR1) gene at the Xq27.3 site (Turner et al., 1996). In most behaviors are linked to the anxiety level of the individual, a affected individuals, this genetic disorder is caused by a meaningful link because physiological studies have noted trinucleotide (CGG) repeat expansion in the 5' untranslated increased sympathetic and decreased parasympathetic activ- (promoter) region of the FMR1 gene. FMR1 encodes the fragile ity and poor coordination between the systems in children X mental retardation protein (FMRP); a 69kDa protein found and adolescents with FXS (Hall et al., 2009). in most adult and fetal tissues, high concentrations are noted Cognitive tests have indicated a specific pattern of in the brain and testes. The expression of FMRP in the brain strengths and weaknesses. FXS individuals exhibit deficits seems to be experience dependent and is produced in the in visuo-spatial tasks, quantitative skills, short-term and soma and near the synapse (Berry-Kravis et al., 2002, 2011). working memory, expressive language skills, sequential pro- FMRP is essential to the shaping of dendritic spines cessing and executive function (Berry-Kravis et al., 2002; (Davidovic et al., 2011). The protein and network of mRNA Cornish et al., 1999; Freund and Reiss, 1991; Hall et al., 2012; targets and interacting proteins contribute to several forms of Kwon et al., 2001; Maes et al., 1994). Relative strengths synaptic plasticity involving learning and memory processes, include receptive language skills, visual memory, acquisition notably induced by activation of type I metabotropic gluta- of factual information, imitation skills and gestalt processing mate receptor (mGluR; Davidovic et al., 2011). Mice lacking (Berry-Kravis et al., 2002). This population also has suscept- FMRP have impaired long-term potentiations in somatosen- ibility to certain other neuropsychiatric disabilities including sory cortex (Li et al., 2002), visual cortex (Wilson and Cox, ASD, ADHD, anxiety disorders, and neurological disorders 2007), olfactory cortex (Larson et al., 2005), cingulate cortex, such as epilepsy (Hessl et al., 2001; Pretorius et al., 1998). and amygdala (Zhao et al., 2005) and enhanced long-term While the genetic cause of FXS has been found, the neurolo- depression in hippocampus (Huber et al., 2002). In synapto- gical basis of FXS symptoms continues to be unknown. MRI somal preparations, stimulation of mGluR results in a FMRP- studies have found that individuals with FXS have enlarged dependent increase in protein synthesis (Weiler et al., 1997, lateral ventricles and increased caudate nucleus volumes 2004). It is hypothesized that a decrease in Fmr1 functionally relative to control subjects (Reiss et al., 1995). Anatomical affects the protein interaction network with direct conse- studies of post-mortem brains have revealed that dendritic quences on signaling cascade and cellular metabolism 208 brain research 1580 (2014) 199–218

(Davidovic et al., 2011). There are two different FMR1 muta- Confirmation of the hypothesis that OT would have tions, full mutation and permutation (Goodrich-Hunsaker beneficial consequences in FXS would be an increase in eye et al., 2011a, 2011b). Premutation, associated with Fragile gaze frequency, a reduction in physiological arousal, and a X-associated tremor/ataxia syndrome (FXTAS; Wang et al., decrease in salivary cortisol. The researchers observed a 2010), has a repeat length of 50–200 and does not usually significant main effect with OT. As compared to placebo, cause mental deficits, but shyness, anxiety, and premature 24IU OT led to a significant increase in eye gaze frequency. ovarian failure have been known to occur. Premutations do They found a significant decrease in salivary cortisol for the appear however to influence translation of FMR1 mRNA (Feng 48 IU dose as compared to placebo. No effects were observed et al., 1995). In many individuals with premutations, excess in the physiological measurements (HR, HRV, and RSA), FMR1 mRNA is produced, yet below normal FMRP is synthe- however, given the small sample and heterogeneous popula- sized (Tassone et al., 2000a, 2000b) and may contribute to tion additional research is needed. Based upon the data to approximately 10% of male and 2–3% of female ASD cases date, Hall and colleagues hypothesized that in FXS adminis- (Wang et al., 2010). Upon female transmission the premuta- tration of OT may dampen amygdala reactivity towards social tion can become a full mutation. FXS is caused by full stimuli that causes anxiety (Kirsch et al., 2005; Petrovic et al., mutation which is 4200 trinucleotide repeats, and results 2008), decrease HPA axis activation, and increase social in hypermethylation of the gene and transcriptional silencing motivation (Witt and Insel, 1992; Witt et al., 1992). (Tassone et al., 2000a). This creates an FMRP deficiency in the brain, which leads to FXS presentation (McLennan et al., 2011; 5.2. Animal models: a way to look at moderators Tassone et al., 2000a). Very rarely have other mutations in the of neurodevelopmental pathways and outcomes? FMR1 gene involving deletions (Gedeon et al., 1992; Wohrle et al., 1992) or a point mutation (De Boulle et al., 1993) Animal models have proven indispensable in the under- resulted in symptoms identical or even more severe standing of diseases and disorders, and in the development than FXS. of pharmaceuticals used to treat them. The quality of an animal model is ascertained based on how well it can meet 5.1. IN-OT as treatment for FXS certain criteria of validity. Three of these criteria are con- struct, face and predictive validity (Bernardet and Crusio, Evidence has supported the investigation of OT as a treat- 2006). How well the model's behavioral traits resemble the ment for FXS (Bartz and Hollander, 2006; Hall et al., 2012; core traits of the disorder is face validity. Predictive validity is Hollander et al., 2007). As described earlier OT, released established when a drug reduces or improves symptoms in endogenously or given exogenously, has been associated both the model and human. Construct validity is the “quality” with positive social behaviors, reductions in anxiety, obses- of the model, its ability to accurately measure or represent siveness and stress reactivity, the central release of AVP and what it claims to be measuring (Cronbach and Meehl, 1955). other peptides such as corticotropin-releasing factor, and There are several FXS animal models, three in mice and a may serve to counter the defensive behavioral strategies drosophila model that meet multiple criteria. associated with stressful experiences (Carter, 2007). However, Two homologous genes to Fmr1 in vertebrates are Fxr1 similar to ASD, available treatments for FXS focus on mana- (fragile X related gene) and Fxr2. The genes are highly ging symptoms: stimulants are prescribed for attention def- homologous at the protein structure level and bind mRNA icit and hyperactivity; SSRI and antipsychotics treat and bind to FMRP. Their proteins, FXR1P and FXR2P are both aggression associated with anxiety; and carbamazepine are expressed in the brain and specifically in cell bodies, but they used for treatment of seizures (Hampson et al., 2011). Cur- are also found in the dendrites near the synapse. Both Fxr1 rently, there are no treatments on the market targeting the and Fxr2 KO mice have been produced. FXR1P deficient / molecular abnormalities of FXS (Gurkan and Hagerman, (Fxr1À À) mice die within 24 h of birth, while heterozygous 2012). Recent studies have begun to investigate IN-OT due mice exhibit abnormal limb musculature. Fxr2 KO have a to the autistic-like behaviors observed in FXS, and the social normal lifespan, learning deficits similar to Fmr1 KO, and and anxiolytic effects of OT. circadian rhythm deficits (Berry-Kravis et al., 2011). The fruit As of the time of publication, very few studies had been fly model, a mutant lacking dFmr1 (also known as dFxr) performed to research the effect of IN-OT on FXS, especially protein, exhibits overextension of neurites during develop- in humans. One such study was conducted by Hall et al. ment of mushroom bodies (brain region linked with memory) (2012). They set up a randomized double-blind placebo-con- and have a behavioral phenotype that includes circadian trolled single-dose trial performed with intranasal adminis- rhythm abnormalities and altered courtship behavior tration of placebo, 24IU OT and 48 IU OT. Studying eight low (Berry-Kravis et al., 2011; Gatto and Broadie, 2009). functioning males between the ages of 13 and 28 years with One of the best characterized animal models of FXS was FXS, they hypothesized that the prosocial and anxiolytic developed in 1994 by the Dutch-Belgian Fragile X Consortium. effects of OT would reduce, if not alleviate, socially inap- This mouse model, which lacks FMRP throughout its lifespan, propriate behaviors and social anxiety. The group collected corresponds to the molecular endpoint of the human disease. eye gaze frequency, heart rate, respiratory sinus arrhythmia, This mouse was created by inserting a neomycin cassette heart rate variability (HRV) during two social challenges into exon 5 of the murine Fmr1 gene. The insert disrupts the (10 min total in length), and salivary cortisol levels before transcription of Fmr1 mRNA causing an absence of FMRP. and after the challenge, which was conducted 50 min after Even though the cause may not be identical, this mouse OT administration. model exhibits behavioral similarities to FXS. Fmr1 KO, in brain research 1580 (2014) 199–218 209

comparison to the wild-type (WT) strain, are described as mouse but not in the human. Third, the behavioral para- having lower than normal levels of initial social interactions digms thus far applied to the mouse model do not efficiently (Mineur et al., 2006), fail to show a preference for social assay or correlate with the cognitive domains most affected novelty, and display inappropriate social responses in individuals with FXS. (Pietropaolo et al., 2011). In contrast, the OT and AVP knock- out (OTKO or AVPKO) mice display high levels of social 5.3. Examining OT in neurodevelopmental animal models: contact that does not diminish over time, but also fail to a way to examine early effects? show indication of familiarity (Crawley et al., 2007). While macroorchidism is observed in the Fmr1 KO animals Below we present preliminary data examining the OT and (Bakker et al., 1994)aswellasinFXSpatients,thebehavioral AVP systems in a mouse model in order to learn more about patterns differ between the patients and the KO. By most FXS pathway interactions during development. The methods accounts Fmr1 KO mice appear to have relatively normal used in this study are described in the text for Fig. 1. These behavior, but research has shown that the behavioral and preliminary results, based upon counts of immune-reactive cognitive deficits of the KO are actually quite subtle and parallel cells, suggest a reduction in both OT-positive (Fig. 1) and AVP- FXS patients (Berry-Kravis et al., 2002; D'Hooge et al., 1997; positive (Fig. 2) cells in the PVN of Fmr1 KO as compared to Paradee et al., 1999; Peier et al., 2000). Among the behavioral WT (Table 2). A trend, although not significant, towards lower phenotypes displayed in Fmr1 KO are deficits in object recogni- OT-positive cells was also noted in the SON (Table 3). To tion memory (including a failure to habituate to objects), and analyze possible differences in the OXTR, the abundance of impairment of spatial memory (Mineur et al., 2002). OXTR-immunoreactive cells were also measured in the hip- Several studies indicate that Fmr1 KO mice are hyperactive pocampus, retrosplenial granular and piriform cortices. None and show indications of increased anxiety (Bakker et al., 1994; of these areas showed a significant difference in OXTR- Mineur et al., 2002; Spencer et al., 2005) and sensory hyperre- immunoreactive cell density as compared to WT mice sponsiveness, especially to auditory stimuli (Chen and Toth, (p40.05). The PVN is an important component of the HPA 2001; Frankland et al., 2004; Nielsen et al., 2002). Loud tones axis, and reductions in OT-positive and AVP-positive cells of may induce audiogenic seizures. Fmr1 KO also exhibit abnor- the PVN might be associated with deficits in the capacity to mal social interactions, including a general reduction in regulate emotional reactivity. Earlier work in voles has social contact and a failure to show social recognition suggested that either OT or AVP may support a general (Bernardet and Crusio, 2006; Mineur et al., 2006; Spencer tendency toward social contact (Cho et al., 1999). Thus, the et al., 2005; Yan et al., 2004). In some tasks there is variability absence of either OT or AVP in the presence of the other did in the results (i.e. complex visual and auditory discriminant not produce an “asocial” animal. However, selective social tasks and activity level in an open field). The symptom preferences, such as those necessary for pair bond formation, variability among mouse models for FXS may be due to appear to require stimulation of both OT and AVP receptors. differences in genetic backgrounds. A similar hypothesis The importance of both OT and AVP to selective behaviors has been proposed to explain the variability observed in the also may be supported by the fact that mice “knocked-out” symptoms of FXS patients. For example, in research by for either OT or the OXTR no longer exhibited selective social Pietropaolo et al. (2011), the validity of the Fmr1 KO mouse memory (Young and Flanagan-Cato, 2012). on the B6 background was tested against WT and Fmr1 KO on Although the preliminary data shown here for Fmr1 KO the FVB background. They found the Fmr1 KO on the B6 mice need to be replicated in a larger sample and in other background to be a good model for FXS and a suitable model animal models, we include these findings as an example of for ASD (Pietropaolo et al., 2011; Yan et al., 2004). possible approaches to examining the role of peptides includ- These mice also have neuropathologic phenotypes that ing, OT and AVP, in molecularly characterized genetic syn- are similar to FXS patients including density of dendritic dromes. Work across these models also could provide spines of pyramidal neurons in the visual and somatosensory additional insight regarding the role of OT and AVP in early cortices that are greater in adult Fmr1 KO than WT. Some development, especially in syndromes in which atypical brain areas, in both mice Fmr1 KO mice and FXS patients, trajectories in social development occur. have spines that appear similar to developing versus mature spines (Berry-Kravis et al., 2002). Absence of FMRP in both humans and mice results in improper development of den- 6. Conclusion and next steps dritic spines on cortical pyramidal neurons (Comery et al., 1997; Irwin et al., 2000; Irwin et al., 2001; Irwin et al., 2002). Each of the disorders described here (ASD, PWS, WS and FXS) The use of the Fmr1 KO mouse has also provided some insight is unique and each condition is characterized by atypical into the normal cellular function of FMRP. The subtle cogni- social behaviors, often with a tendency toward high levels of tive deficits of Fmr1 KO mice present difficulties for preclinical anxiety. Given the importance of OT and AVP to mammalian testing of potential treatments, and highlight how complex social behaviors and anxiety, the neuropeptides' investigative the relationship between the mouse and human phenotypes value in these syndromes is not unexpected. This review are. One possibility is that the cognitive processes in which summarized the possible role of OT in these NDD (Table 4) FMRP plays a vital role in humans are poorly developed in through experiments conducted by others and ourselves. mice; thus mice lacking FMRP are not particularly disabled, at Each of these early developmental disorders displays least compared to severely-affected patients. A second pos- alterations in the OT system. These changes may impact sibility is that other proteins can compensate for FMRP in the behavior and emotional regulation through a variety of 210 brain research 1580 (2014) 199–218

molecular and neuroendocrine pathways. For example, our At present, the largest concentration of studies on the preliminary data suggests a decreased number of OT-positive role of dysregulated OT pathways has been conducted in and AVP-positive cells in the PVN of Fmr1 KO mice, a mouse ASD. However, as new data are emerging it is striking that model for FXS. Individuals with PWS have shown lower levels other disorders with phenotypes marked by abnormal of OT in CSF (Martin et al., 1998) and fewer OT producing cells social behavior, as well as anxiety (in some cases mani- in the PVN (Swaab et al., 1995). A subgroup of ASD affected fested by RRB) also appear to have abnormalities found children also appeared to have lower plasma OT levels within the OT system. For example, as in ASD, individuals (Modahl et al., 1998). In contrast WS, which is characterized with PWS have difficulty with social competence by hypersociability, had a positive correlation between OT (Dimitropoulos et al., 2013), are aloof and avoid eye contact levels and increased stranger approach and decreased adap- (Dimitropoulos et al., 2009). Furthermore, RRB is also evi- tive social behavior (Dai et al., 2012). denced in PWS (Greaves et al., 2006), although to a lesser

Wild Type Knockout

200µm

50µm

Fig. 1 – Expression of OT in the paraventricular nucleus (PVN), as measured by ICC, is reduced in Fmr1 KO mice as compared to the wild-type (WT). Methods: The animal model was generated with WT and Fmr1 KO mice from a colony founded with stock obtained from the Jackson Laboratory (Bar Harbor, ME, USA) that was backcrossed onto a B6 background 410 generations. Mice were genotyped using primers described previously (the Dutch-Belgian Fragile-X Consortium, 1994). Cells were stained using the immunocytochemical (ICC) staining procedures, following protocols described in early work on OT and AVP in voles (Yamamoto et al., 2004). All sections were double-stained for NeuN (a marker that stains cell nuclei only in neurons), which allowed precise localization of cytoarchitectonic boundaries. Stained sections were mounted on subbed slides and examined with OT and AVP antibodies (OT antibodies were generously provided by M. Morris and AVP antibodies were obtained from MP Biomedical #647171, formerly ICN; Solon, OH, USA). Slices of tissue for each animal were categorized as described in (Paxinos and Franklin, 2004) and carefully matched across subjects to allow comparable sections. Imaged slides were captured at 10 , then coded and scored by an experimentally blind scorer using Image J (NIH, Bethesda, MD) software. OT Â and AVP stained cells in the PVN of the hypothalamus regions were stained separately for OT and AVP (N 6–7 mice per ¼ group). Boxed sampling areas were: 125 125 μm2 (PVN total staining density), 250 375 μm2 (PVN fibers), 93.75 93.75 μm2 Â Â Â for cell counts bilaterally in both the PVN and SON. brain research 1580 (2014) 199–218 211

Wild Type Knockout

200µm

50µm

Fig. 2 – Expression of AVP in the paraventricular nucleus (PVN). In Fmr1 KO mice, as compared to the wild-type (WT) AVP expression is reduced as measured by ICC.

Table 2 – Number of OT and AVP-positive cells in PVN of Fmr1 KO versus WT mice.

OT AVP

WT Knockout WT Knockout

PVN 17†72n 973 p 0.047 1072472 p 0.05 ¼ ¼ † Means number of positive cell/0.2 mm2. n Means 7SE (N 6–7/group). ¼

Table 3 – Number of OT and AVP-positive cells in SON of Fmr1 KO versus WT mice.

OT AVP

WT Knockout WT Knockout

SON 13†72n 873 p 0.254 1172672 p 0.107 ¼ ¼ † Means number of positive cell/0.08 mm2. n Means 7SE (N 6–7/group). ¼

degree than in ASD as measured by the Repetitive Behavior WS and ASD also share commonalities as both are marked Scale-Revised (RBS-R; Flores et al., 2011). A subset of the by abnormal social phenotypes and anxiety. However, unlike genetic region associated with PWS is also associated with PWS, individuals with WS show a phenotype that is markedly an increased risk for ASD, as maternally inherited duplica- different from ASD. Although both groups are at risk for tions of the 15q11-13 region are associated with 1–3% of anxiety, individuals with ASD show higher levels of social ASD cases (Bolton et al., 2001; Cook et al., 1997; Vorstman phobia and separation anxiety, as well as higher rates of RRB. et al., 2006). However, individuals with WS have higher scores on 212 brain research 1580 (2014) 199–218

Table 4 – A Summary of OT Affects on NDD.

Disorder Neuropeptide system affected References

Autism spectrum Atypical (↓↑ observed) levels of OT in blood (human) Modahl et al. (1998) disorders IN-OT ↑ social task performance and ↓ repetitive behaviors Andari et al. (2010) and Hollander et al. (2003) (human) To be studied: human neuropathology and animal models For a more extensive list of human trials see Table 1

Prader–Willi syndrome ↓ OT producing cells in the PVN (human) Swaab et al. (1995) ↓ Level of OT in CSF (human) Martin et al. (1998) IN-OT ↑ trust and ↓ disruptive behaviors (human) Tauber et al. (2011) ↓ Hypothalamic OT in Maged1 deletion model (animal) Dombret et al. (2012)

Williams syndrome ↑ OT levels (human) Dai et al. (2012) To be studied: human neuropathology and animal models

Fragile X syndrome ↓ OT and AVP cells in the PVN (Fmr1 KO mice) See Tables 2 and 3, and Figs. 1 and 2 þ þ IN-OT ↑ eye gaze frequency (human) Hall et al. (2009) and Hall et al. (2012) IN-OT ↓ salivary cortisol (human) Hall et al. (2012) To be studied: human neuropathology

measures of generalized anxiety (Rodgers et al., 2012). WS is characterized by an increase in OT levels (Dai et al., 2012), as Acknowledgments well as a deletion of the 7q11.23 region, as opposed to a de novo duplication which leads to ASD (Sanders et al., 2011). It This work was supported in part by NIH K23MH082121 (SJ). is likely that features of ASD and WS are manifestations of The authors would like to acknowledge Dr. John Larson for gene dosage effects on similar behaviors. Studies of FXS and his contribution of the Fmr1 KO mice. We would also like to ASD mechanisms may also inform each other, as mutations thank Jeanine Leary and Jennifer Speak for their assistance in in mGluR5 can contribute to the diagnosis of FXS or ASD, and manuscript formatting and preparation. mGluR5 antagonists have shown promise in alleviating ASD symptoms in mouse models (Silverman et al., 2012) as well as references FXS pathology. Due to the rarity of these disorders and the complex animal models needed to study them, many of these experiments have small sample sizes. However, these studies Abramson, R.K., Wright, H.H., Carpenter, R., Brennan, W., remain significant and together provide a motivation and Lumpuy, O., Cole, E., Young, S.R., 1989. Elevated blood serotonin in autistic probands and their first-degree relatives. direction for future research in NDD, especially disorders J. Autism Dev. Disord. 19, 397–407. with dysfunctional social behaviors as a symptom. American Psychiatric Association, 2013. In: Diagnostic and As summarized in this review, the dysregulation of the OT Statistical Manual of Mental Disorders 5th ed. American system in animals and humans is associated with marked Psychiatric Publishing, Arlington, VA. deficits in social behavior as well as anxiety. This common- Anagnostou, E., Soorya, L., Chaplin, W., Bartz, J., Halpern, D., ality across multiple NDD may indicate a shared OT pathway Wasserman, S., Wang, A.T., Pepa, L., Tanel, N., Kushki, A., that is affected during development. The use of animal Hollander, E., 2012. Intranasal oxytocin versus placebo in the treatment of adults with autism spectrum disorders: a models, particularly those developed for FXS, WS and PWS randomized controlled trial. Mol. Autism 3, 16. will provide insight into such a pathway, as these disorders Andari, E., Duhamel, J.R., Zalla, T., Herbrecht, E., Leboyer, M., have well characterized genetics. In contrast, there are over a Sirigu, A., 2010. Promoting social behavior with oxytocin in 103 disease genes and 44 genomic loci reported to be involved high-functioning autism spectrum disorders. Proc. Natl. Acad. in ASD (Betancur, 2011). However, unlike ASD research, there Sci. USA 107, 4389–4394. is a lack of human data on the pathophysiology of FXS, WS Appenrodt, E., Schnabel, R., Schwarzberg, H., 1998. Vasopressin and PWS, and pharmacological interventions. Ideally, scien- administration modulates anxiety-related behavior in rats. tists want to identify specific molecular pathways to target Physiol. Behav. 64, 543–547. Bagdy, G., Kalogeras, K.T., 1993. Stimulation of 5-HT1A and 5-HT2/ distinct syndromes and disorders for treatment. However, 5-HT1C receptors induce oxytocin release in the male rat. many effective medical treatments, such as drugs for hyper- Brain Res. 611, 330–332. tension, modulate common neurochemical or hormone path- Bailey, A., Le Couteur, A., Gottesman, I., Bolton, P., Simonoff, E., ways that are downstream from etiologically contributing Yuzda, E., Rutter, M., 1995. Autism as a strongly genetic factors. Combining the strengths of human and animal disorder: evidence from a British twin study. Psychol. Med. 25, model studies across these NDD may provide important clues 63–77. into the developmental role of OT. Additionally, as general Bakker, C.E., Verheij, C., Willemsen, R., Vanderhelm, R., Oerlemans, F., Vermey, M., Bygrave, A., Hoogeveen, A.T., mechanisms underlying social and emotional behaviors are Oostra, B.A., Reyniers, E., Deboulle, K., Dhooge, R., Cras, P., fi speci ed, it may become possible to elucidate the complex Vanvelzen, D., Nagels, G., Martin, J.J., Dedeyn, P.P., Darby, J.K., neurophysiology of and create treatment targets for FXS, Willems, P.J., 1994. Fmr1 knockout mice—a model to study PWS, WS and ASD. fragile-X mental-retardation. Cell 78, 23–33. brain research 1580 (2014) 199–218 213

Bartz, J.A., Hollander, E., 2006. The neuroscience of affiliation: Chini, B., Mouillac, B., Balestre, M.N., Trumpp-Kallmeyer, S., forging links between basic and clinical research on Hoflack, J., Hibert, M., Andriolo, M., Pupier, S., Jard, S., Barberis, neuropeptides and social behavior. Horm. Behav. 50, 518–528. C., 1996. Two aromatic residues regulate the response of the Baumgartner, T., Heinrichs, M., Vonlanthen, A., Fischbacher, U., human oxytocin receptor to the partial agonist arginine Fehr, E., 2008. Oxytocin shapes the neural circuitry of trust vasopressin. FEBS Lett. 397, 201–206. and trust adaptation in humans. Neuron 58, 639–650. Cho, M.M., DeVries, A.C., Williams, J.R., Carter, C.S., 1999. The Bernardet, M., Crusio, W.E., 2006. Fmr1 KO mice as a possible effects of oxytocin and vasopressin on partner preferences in model of autistic features. Sci. World J. 6, 1164–1176. male and female prairie voles (Microtus ochrogaster). Behav. Berry-Kravis, E., Grossman, A.W., Crnicz, L.S., Greenough, W.T., Neurosci. 113, 1071–1079. 2002. Understanding fragile X syndrome. Curr. Paediatr. 12, Chugani, D.C., Muzik, O., Behen, M., Rothermel, R., Janisse, J.J., 316–324. Lee, J., Chugani, H.T., 1999. Developmental changes in brain Berry-Kravis, E., Knox, A., Hervey, C., 2011. Targeted treatments serotonin synthesis capacity in autistic and nonautistic for fragile X syndrome. J. Neurodev. Disord. 3, 193–210. children. Ann. Neurol. 45, 287–295. Betancur, C., 2011. Etiological heterogeneity in autism spectrum Comery, T.A., Harris, J.B., Willems, P.J., Oostra, B.A., Irwin, S.A., disorders: more than 100 genetic and genomic disorders and Weiler, I.J., Greenough, W.T., 1997. Abnormal dendritic spines still counting. Brain Res. 1380, 42–77. in fragile X knockout mice: maturation and pruning deficits. Bittel, D.C., Kibiryeva, N., McNulty, S.G., Driscoll, D.J., Butler, M.G., Proc. Natl. Acad. Sci. USA 94, 5401–5404. White, R.A., 2007a. Whole genome microarray analysis of gene Cook Jr., E.H., Lindgren, V., Leventhal, B.L., Courchesne, R., expression in an imprinting center deletion mouse model of Lincoln, A., Shulman, C., Lord, C., Courchesne, E., 1997. Prader–Willi syndrome. Am. J. Med. Genet. A 143, 422–429. Autism or atypical autism in maternally but not paternally Bittel, D.C., Kibiryeva, N., Sell, S.M., Strong, T.V., Butler, M.G., derived proximal 15q duplication. Am. J. Hum. Genet. 60, 2007b. Whole genome microarray analysis of gene expression 928–934. in Prader–Willi syndrome. Am. J. Med. Genet. A 143, 430–442. Cornish, K.M., Munir, F., Cross, G., 1999. Spatial cognition in males Boccaccio, I., Glatt-Deeley, H., Watrin, F., Roeckel, N., Lalande, M., with Fragile-X syndrome: evidence for a neuropsychological Muscatelli, F., 1999. The human MAGEL2 gene and its mouse phenotype. Cortex 35, 263–271. Crawley, J.N., Chen, T., Puri, A., Washburn, R., Sullivan, T.L., Hill, homologue are paternally expressed and mapped to the J.M., Young, N.B., Nadler, J.J., Moy, S.S., Young, L.J., Caldwell, Prader–Willi region. Hum. Mol. Genet. 8, 2497–2505. H.K., Young, W.S., 2007. Social approach behaviors in oxytocin Bolton, P.F., Dennis, N.R., Browne, C.E., Thomas, N.S., Veltman, M. knockout mice: comparison of two independent lines tested in W., Thompson, R.J., Jacobs, P., 2001. The phenotypic different laboratory environments. Neuropeptides 41, 145–163. manifestations of interstitial duplications of proximal 15q Cronbach, L.J., Meehl, P.E., 1955. Construct validity in with special reference to the autistic spectrum disorders. Am. psychological tests. Psychol. Bull. 52, 281–302. J. Med. Genet. 105, 675–685. D’Hooge, R., Nagels, G., Franck, F., Bakker, C.E., Reyniers, E., Borghgraef, M., Fryns, J.P., Van Den Berghe, H., 1990. Psychological Storm, K., Kooy, R.F., Oostra, B.A., Willems, P.J., De Deyn, P.P., profile and behavioural characteristics in 12 patients with 1997. Mildly impaired water maze performance in male Fmr1 Prader–Willi syndrome. Genet. Couns. 1, 141–150. knockout mice. Neuroscience 76, 367–376. Born, J., Lange, T., Kern, W., McGregor, G.P., Bickel, U., Fehm, H.L., Dadds, M.R., Macdonald, E., Cauchi, A., Williams, K., Levy, F., 2002. Sniffing neuropeptides: a transnasal approach to the Brennan, J., Nasal oxytocin for social deficits in childhood human brain. Nat. Neurosci. 5, 514–516. autism: a randomized controlled trial, J. Autism Dev. Disord. Brodkin, E.S., 2007. BALB/c mice: low sociability and other in press, [Epub ahead of print]. phenotypes that may be relevant to autism. Behav. Brain Res. Dai, L., Carter, C.S., Ying, J., Bellugi, U., Pournajafi-Nazarloo, H., 176, 53–65. Korenberg, J.R., 2012. Oxytocin and vasopressin are Brown, W.T., 1990. The fragile X: progress toward solving the dysregulated in Williams syndrome, a genetic disorder puzzle. Am. J. Hum. Genet. 47, 175–180. affecting social behavior. PLoS One 7, e38513. Campbell, D.B., Datta, D., Jones, S.T., Batey Lee, E., Sutcliffe, J.S., Davidovic, L., Navratil, V., Bonaccorso, C.M., Catania, M.V., Hammock, E.A., Levitt, P., 2011. Association of oxytocin Bardoni, B., Dumas, M.E., 2011. A metabolomic and systems receptor (OXTR) gene variants with multiple phenotype biology perspective on the brain of the fragile X syndrome domains of autism spectrum disorder. J. Neurodev. Disord. 3, mouse model. Genome Res. 21, 2190–2202. 101–112. De Boulle, K., Verkerk, A.J., Reyniers, E., Vits, L., Hendrickx, J., Van Carter, C.S., 1998. Neuroendocrine perspectives on social Roy, B., Van den Bos, F., de Graaff, E., Oostra, B.A., Willems, P.J., attachment and love. Psychoneuroendocrinology 23, 779–818. 1993. A point mutation in the FMR-1 gene associated with Carter, C.S., 2007. Sex differences in oxytocin and vasopressin: fragile X mental retardation. Nat. Genet. 3, 31–35. implications for autism spectrum disorders?. Behav. Brain Dimitropoulos, A., Feurer, I.D., Roof, E., Stone, W., Butler, M.G., Res. 176, 170–186. Sutcliffe, J., Thompson, T., 2000. Appetitive behavior, Cascio, C.J., Foss-Feig, J.H., Heacock, J.L., Newsom, C.R., Cowan, R. compulsivity, and neurochemistry in Prader–Willi syndrome. L., Benningfield, M.M., Rogers, B.P., Cao, A., 2012. Response of Ment. Retard. Dev. Disabil. Res. Rev. 6, 125–130. neural reward regions to food cues in autism spectrum Dimitropoulos, A., Ho, A.Y., Klaiman, C., Koenig, K., Schultz, R.T., disorders. J. Neurodev. Disord. 4, 9. 2009. A comparison of behavioral and emotional Cassidy, S.B., Schwartz, S., Miller, J.L., Driscoll, D.J., 2011. Prader– characteristics in children with autism, Prader–Willi Willi syndrome. Genet. Med.. syndrome, and Williams syndrome. J. Ment. Health Res. Cassidy, S.B., Schwartz, S., Miller, J.L., Driscoll, D.J., 2012. Prader– Intellect. Disabil. 2, 220–243. Willi syndrome. Genet. Med. 14, 10–26. Dimitropoulos, A., Ho, A., Feldman, B., 2013. Social Chaste, P., Leboyer, M., 2012. Autism risk factors: genes, responsiveness and competence in Prader–Willi syndrome: environment, and gene-environment interactions. Dialogues direct comparison to autism spectrum disorder. J. Autism Dev. Clin. Neurosci. 14, 281–292. Disord. 43, 103–113. Chen, L., Toth, M., 2001. Fragile X mice develop sensory Dombret, C., Nguyen, T., Schakman, O., Michaud, J.L., Hardin- hyperreactivity to auditory stimuli. Neuroscience 103, Pouzet, H., Bertrand, M.J., De Backer, O., 2012. Loss of Maged1 1043–1050. results in obesity, deficits of social interactions, impaired 214 brain research 1580 (2014) 199–218

sexual behavior and severe alteration of mature oxytocin Gosch, A., Pankau, R., 1994. Social–emotional and behavioral production in the hypothalamus. Hum. Mol. Genet. 21, adjustment in children with Williams–Beuren syndrome. Am. 4703–4717. J. Med. Genet. 53, 335–339. Domes, G., Heinrichs, M., Michel, A., Berger, C., Herpertz, S., 2007. Gosch, A., Pankau, R., 1997. Personality characteristics and Oxytocin improves “Mind-Reading” in humans. Biol. behaviour problems in individuals of different ages with Psychiatry 61, 731–733. Williams syndrome. Dev. Med. Child Neurol. 39, 527–533. Domes, G., Sibold, M., Schulze, L., Lischke, A., Herpertz, S.C., Greaves, N., Prince, E., Evans, D.W., Charman, T., 2006. Repetitive Heinrichs, M., 2013. Intranasal oxytocin increases covert and ritualistic behaviour in children with Prader–Willi attention to positive social cues. Psychol. Med. 43, 1747–1753. syndrome and children with autism. J. Intellect. Disabil. Res. Ebstein, R.P., Israel, S., Lerer, E., Uzefovsky, F., Shalev, I., Gritsenko, I., 50, 92–100. Riebold, M., Salomon, S., Yirmiya, N., 2009. Arginine vasopressin Green, L., Fein, D., Modahl, C., Feinstein, C., Waterhouse, L., and oxytocin modulate human social behavior. Ann. N. Y. Acad. Morris, M., 2001. Oxytocin and autistic disorder: alterations in Sci. 1167, 87–102. peptide forms. Biol. Psychiatry 50, 609–613. Gross-Tsur, V., Landau, Y.E., Benarroch, F., Wertman-Elad, R., Ebstein, R.P., Knafo, A., Mankuta, D., Chew, S.H., Lai, P.S., 2012. Shalev, R.S., 2001. Cognition, attention, and behavior in The contributions of oxytocin and vasopressin pathway genes Prader–Willi syndrome. J. Child Neurol. 16, 288–290. to human behavior. Horm. Behav. 61, 359–379. Guastella, A.J., Mitchell, P.B., Dadds, M.R., 2008. Oxytocin Einfeld, S.L., Tonge, B.J., Florio, T., 1997. Behavioral and emotional increases gaze to the eye region of human faces. Biol. disturbance in individuals with Williams syndrome. Am. J. Psychiatry 63, 3–5. Ment. Retard. 102, 45–53. Guastella, A.J., Einfeld, S.L., Gray, K.M., Rinehart, N.J., Tonge, B.J., Ewart, A.K., Morris, C.A., Atkinson, D., Jin, W., Sternes, K., Lambert, T.J., Hickie, I.B., 2010. Intranasal oxytocin improves Spallone, P., Stock, A.D., Leppert, M., Keating, M.T., 1993. emotion recognition for youth with autism spectrum Hemizygosity at the elastin locus in a developmental disorder, disorders. Biol. Psychiatry 67, 692–694. Williams syndrome. Nat. Genet. 5, 11–16. Gurkan, C.K., Hagerman, R.J., 2012. Targeted treatments in autism Feldman, R., Zagoory-Sharon, O., Weisman, O., Schneiderman, I., and Fragile X syndrome. Res. Autism Spectr. Disord. 6, Gordon, I., Maoz, R., Shalev, I., Ebstein, R.P., 2012. Sensitive 1311–1320. parenting is associated with plasma oxytocin and Hagerman, R., Hoem, G., Hagerman, P., 2010. Fragile X and polymorphisms in the OXTR and CD38 genes. Biol. Psychiatry autism: Intertwined at the molecular level leading to targeted 72, 175–181. treatments. Mol. Autism 1, 12. Feng, Y., Zhang, F., Lokey, L.K., Chastain, J.L., Lakkis, L., Eberhart, D., Hagerman, R.J., Berry-Kravis, E., Kaufmann, W.E., Ono, M.Y., Warren, S.T., 1995. Translational suppression by trinucleotide Tartaglia, N., Lachiewicz, A., Kronk, R., Delahunty, C., Hessl, repeat expansion at FMR1. Science 268, 731–734. D., Visootsak, J., Picker, J., Gane, L., Tranfaglia, M., 2009. Ferguson, J.N., Young, L.J., Hearn, E.F., Matzuk, M.M., Insel, T.R., Advances in the treatment of fragile X syndrome. Pediatrics Winslow, J.T., 2000. Social amnesia in mice lacking the 123, 378–390. oxytocin gene. Nat. Genet. 25, 284–288. Hall, S.S., 2009. Treatments for fragile X syndrome: a closer look Flores, C.G., Valcante, G., Guter, S., Zaytoun, A., Wray, E., Bell, L., at the data. Dev. Disabil. Res. Rev. 15, 353–360. Jacob, S., Lewis, M.H., Driscoll, D.J., Cook Jr., E.H., Kim, S.J., Hall, S.S., Lightbody, A.A., Huffman, L.C., Lazzeroni, L.C., Reiss, A. 2011. Repetitive behavior profiles: consistency across autism L., 2009. Physiological correlates of social avoidance behavior spectrum disorder cohorts and divergence from Prader–Willi in children and adolescents with fragile x syndrome. J. Am. syndrome. J. Neurodev. Disord. 3, 316–324. Acad. Child Adolesc. Psychiatry 48, 320–329. Frankland, P.W., Wang, Y., Rosner, B., Shimizu, T., Balleine, B.W., Hall, S.S., Lightbody, A.A., McCarthy, B.E., Parker, K.J., Reiss, A.L., Dykens, E.M., Ornitz, E.M., Silva, A.J., 2004. Sensorimotor 2012. Effects of intranasal oxytocin on social anxiety in males gating abnormalities in young males with fragile X syndrome with fragile X syndrome. Psychoneuroendocrinology 37, and Fmr1-knockout mice. Mol. Psychiatry 9, 417–425. 509–518. Freund, L.S., Reiss, A.L., 1991. Cognitive profiles associated with Hammock, E., Veenstra-VanderWeele, J., Yan, Z., Kerr, T.M., the fra(X) syndrome in males and females. Am. J. Med. Genet. Morris, M., Anderson, G.M., Carter, C.S., Cook, E.H., Jacob, S., 2012. Examining autism spectrum disorders by biomarkers: 38, 542–547. example from the oxytocin and serotonin systems. J. Am. Gagliardi, C., Frigerio, E., Burt, D.M., Cazzaniga, I., Perrett, D.I., Acad. Child Adolesc. Psychiatry 51, 712–721 e1. Borgatti, R., 2003. Facial expression recognition in Williams Hampson, D.R., Adusei, D.C., Pacey, L.K., 2011. The neurochemical syndrome. Neuropsychologia 41, 733–738. basis for the treatment of autism spectrum disorders and Gatto, C.L., Broadie, K., 2009. Temporal requirements of the fragile Fragile X syndrome. Biochem. Pharmacol. 81, 1078–1086. x mental retardation protein in modulating circadian clock Hatton, D.D., Sideris, J., Skinner, M., Mankowski, J., Bailey Jr., D.B., circuit synaptic architecture. Front. Neural Circuits 3, 8. Roberts, J., Mirrett, P., 2006. Autistic behavior in children with Gedeon, A.K., Baker, E., Robinson, H., Partington, M.W., Gross, B., fragile X syndrome: prevalence, stability, and the impact of Manca, A., Korn, B., Poustka, A., Yu, S., Sutherland, G.R., et al., FMRP. Am. J. Med. Genet. A 140A, 1804–1813. 1992. Fragile X syndrome without CCG amplification has an Hessl, D., Dyer-Friedman, J., Glaser, B., Wisbeck, J., Barajas, R.G., FMR1 deletion. Nat. Genet. 1, 341–344. Taylor, A., Reiss, A.L., 2001. The influence of environmental Goodrich-Hunsaker, N.J., Wong, L.M., McLennan, Y., Srivastava, S., and genetic factors on behavior problems and autistic Tassone, F., Harvey, D., Rivera, S.M., Simon, T.J., 2011a. Young symptoms in boys and girls with Fragile X syndrome. adult female fragile X premutation carriers show age- and Pediatrics 108, E88. genetically-modulated cognitive impairments. Brain Cognit. Hinton, V.J., Brown, W.T., Wisniewski, K., Rudelli, R.D., 1991. 75, 255–260. Analysis of neocortex in three males with the Fragile X Goodrich-Hunsaker, N.J., Wong, L.M., McLennan, Y., Tassone, F., syndrome. Am. J. Med. Genet. 41, 289–294. Harvey, D., Rivera, S.M., Simon, T.J., 2011b. Adult female Hollander, E., Novotny, S., Hanratty, M., Yaffe, R., DeCaria, C.M., Fragile X premutation carriers exhibit age- and CGG repeat Aronowitz, B.R., Mosovich, S., 2003. Oxytocin infusion reduces length-related impairments on an attentionally based repetitive behaviors in adults with autistic and Asperger’s enumeration task. Front. Hum. Neurosci. 5, 63. disorders. Neuropsychopharmacology 28, 193–198. brain research 1580 (2014) 199–218 215

Hollander, E., Bartz, J., Chaplin, W., Phillips, A., Sumner, J., Soorya, responses to exposure to an infant in male prairie voles. J. L., Anagnostou, E., Wasserman, S., 2007. Oxytocin increases Neuroendocrinol. 24, 874–886. retention of social cognition in autism. Biol. Psychiatry 61, Kirsch, P., 2005. Oxytocin modulates neural circuitry for social 498–503. cognition and fear in humans. J. Neurosci. 25, 11489–11493. Huber, K.M., Gallagher, S.M., Warren, S.T., Bear, M.F., 2002. Altered Kirsch, P., Esslinger, C., Chen, Q., Mier, D., Lis, S., Siddhanti, S., synaptic plasticity in a mouse model of fragile X mental Gruppe, H., Mattay, V.S., Gallhofer, B., Meyer-Lindenberg, A., retardation. Proc. Natl. Acad. Sci. USA 99, 7746–7750. 2005. Oxytocin modulates neural circuitry for social cognition Irwin, S.A., Galvez, R., Greenough, W.T., 2000. Dendritic spine and fear in humans. J. Neurosci. 25, 11489–11493. structural anomalies in Fragile-X mental retardation Korenberg, J.R., Chen, X.N., Hirota, H., Lai, Z., Bellugi, U., Burian, syndrome. Cereb. Cortex 10, 1038–1044. D., Roe, B., Matsuoka, R., 2000. VI. Genome structure and Irwin, S.A., Patel, B., Idupulapati, M., Harris, J.B., Crisostomo, R.A., cognitive map of Williams syndrome. J. Cogn. Neurosci. 12 Larsen, B.P., Kooy, F., Willems, P.J., Cras, P., Kozlowski, P.B., (Suppl. 1), S89–S107. Swain, R.A., Weiler, I.J., Greenough, W.T., 2001. Abnormal Kosfeld, M., Heinrichs, M., Zak, P.J., Fischbacher, U., Fehr, E., 2005. dendritic spine characteristics in the temporal and visual Oxytocin increases trust in humans. Nature 435, 673–676. cortices of patients with Fragile-X syndrome: a quantitative Kuperman, S., Beeghly, J.H., Burns, T.L., Tsai, L.Y., 1985. Serotonin examination. Am. J. Med. Genet. 98, 161–167. relationships of autistic probands and their first-degree Irwin, S.A., Idupulapati, M., Gilbert, M.E., Harris, J.B., Chakravarti, relatives. J. Am. Acad. Child Psychiatry 24, 186–190. A.B., Rogers, E.J., Crisostomo, R.A., Larsen, B.P., Mehta, A., Kwon, H., Menon, V., Eliez, S., Warsofsky, I.S., White, C.D., Dyer- Alcantara, C.J., Patel, B., Swain, R.A., Weiler, I.J., Oostra, B.A., Friedman, J., Taylor, A.K., Glover, G.H., Reiss, A.L., 2001. Greenough, W.T., 2002. Dendritic spine and dendritic field Functional neuroanatomy of visuospatial working memory in characteristics of layer V pyramidal neurons in the visual fragile X syndrome: relation to behavioral and molecular cortex of Fragile-X knockout mice. Am. J. Med. Genet. 111, measures. Am. J. Psychiatry 158, 1040–1051. 140–146. Landgraf, R., Wigger, A., 2003. Born to be anxious: neuroendocrine Ishak, W.W., Kahloon, M., Fakhry, H., 2011. Oxytocin role in and genetic correlates of trait anxiety in HAB rats. Stress 6, enhancing well-being: a literature review. J. Affect. Disord. 130, 111–119. Larson, J., Jessen, R.E., Kim, D., Fine, A.K., du Hoffmann, J., 2005. 1–9. Age-dependent and selective impairment of long-term Jacob, S., Brune, C.W., Carter, C.S., Leventhal, B.L., Lord, C., Cook potentiation in the anterior piriform cortex of mice lacking Jr., E.H., 2007. Association of the oxytocin receptor gene the fragile X mental retardation protein. J. Neurosci. 25, (OXTR) in Caucasian children and adolescents with autism. 9460–9469. Neurosci. Lett. 417, 6–9. Lavi-Itzkovitz, A., Tcherpakov, M., Levy, Z., Itzkovitz, S., Jansen, L.M., Gispen-de Wied, C.C., Wiegant, V.M., Westenberg, H. Muscatelli, F., Fainzilber, M., 2012. Functional consequences of G., Lahuis, B.E., van Engeland, H., 2006. Autonomic and necdin nucleocytoplasmic localization. PLoS One 7, e33786. neuroendocrine responses to a psychosocial stressor in adults Leboyer, M., Philippe, A., Bouvard, M., Guilloud-Bataille, M., with autistic spectrum disorder. J. Autism Dev. Disord. 36, Bondoux, D., Tabuteau, F., Feingold, J., Mouren-Simeoni, M.C., 891–899. Launay, J.M., 1999. Whole blood serotonin and plasma beta- Jarvinen-Pasley, A., Bellugi, U., Reilly, J., Mills, D.L., Galaburda, A., endorphin in autistic probands and their first-degree relatives. Reiss, A.L., Korenberg, J.R., 2008. Defining the social phenotype Biol. Psychiatry 45, 158–163. in Williams syndrome: a model for linking gene, the brain, Leekam, S., BaronCohen, S., Perrett, D., Milders, M., Brown, S., and behavior. Dev. Psychopathol. 20, 1–35. 1997. Eye-direction detection: a dissociation between Jarvinen-Pasley, A., Adolphs, R., Yam, A., Hill, K.J., Grichanik, M., geometric and joint attention skills in autism. Br. J. Dev. Reilly, J., Mills, D., Reiss, A.L., Korenberg, J.R., Bellugi, U., 2010. Psychol. 15, 77–95. Affiliative behavior in Williams syndrome: social perception Lerer, E., Levi, S., Israel, S., Yaari, M., Nemanov, L., Mankuta, D., and real-life social behavior. Neuropsychologia 48, 2110–2119. Nurit, Y., Ebstein, R.P., 2010. Low CD38 expression in Jarvinen, A., Korenberg, J.R., Bellugi, U., 2013. The social lymphoblastoid cells and haplotypes are both associated with phenotype of Williams syndrome. Curr. Opin. Neurobiol. 23, autism in a family-based study. Autism Res. 3, 293–302. 414–422. Leventhal, B.L., Cook Jr., E.H., Morford, M., Ravitz, A., Freedman, Jaselskis, C.A., Cook Jr., E.H., Fletcher, K.E., Leventhal, B.L., 1992. D.X., 1990. Relationships of whole blood serotonin and plasma Clonidine treatment of hyperactive and impulsive children norepinephrine within families. J. Autism Dev. Disord. 20, with autistic disorder. J. Clin. Psychopharmacol. 12, 322–327. 499–511. Jemel, B., Mottron, L., Dawson, M., 2006. Impaired face processing Li, J., Pelletier, M.R., Perez Velazquez, J.L., Carlen, P.L., 2002. in autism: fact or artifact?. J. Autism Dev. Disord. 36, 91–106. Reduced cortical synaptic plasticity and GluR1 expression Jin, D., Liu, H.X., Hirai, H., Torashima, T., Nagai, T., Lopatina, O., associated with fragile X mental retardation protein Shnayder, N.A., Yamada, K., Noda, M., Seike, T., Fujita, K., deficiency. Mol. Cell Neurosci. 19, 138–151. Takasawa, S., Yokoyama, S., Koizumi, K., Shiraishi, Y., Tanaka, Lim, M.M., Wang, Z., Olazabal, D.E., Ren, X., Terwilliger, E.F., S., Hashii, M., Yoshihara, T., Higashida, K., Islam, M.S., Young, L.J., 2004. Enhanced partner preference in a Yamada, N., Hayashi, K., Noguchi, N., Kato, I., Okamoto, H., promiscuous species by manipulating the expression of a Matsushima, A., Salmina, A., Munesue, T., Shimizu, N., single gene. Nature 429, 754–757. Mochida, S., Asano, M., Higashida, H., 2007. CD38 is critical for Lim, M.M., Bielsky, I.F., Young, L.J., 2005. Neuropeptides and the social behaviour by regulating oxytocin secretion. Nature 446, social brain: potential rodent models of autism. Int. J. Dev. 41–45. Neurosci. 23, 235–243. Jorgensen, H., Riis, M., Knigge, U., Kjaer, A., Warberg, J., 2003. Lord, C., Rutter, M., Le Couteur, A., 1994. Autism diagnostic Serotonin receptors involved in vasopressin and oxytocin interview-revised: a revised version of a diagnostic interview secretion. J. Neuroendocrinol. 15, 242–249. for caregivers of individuals with possible pervasive Kanner, L., 1943. Autistic disturbances of affective contact. Nerv. developmental disorders. J. Autism Dev. Disord. 24, 659–685. Child 2, 217–250. Lord, C., Risi, S., Lambrecht, L., Cook Jr., E.H., Leventhal, B.L., Kenkel, W.M., Paredes, J., Yee, J.R., Pournajafi-Nazarloo, H., Bales, DiLavore, P.C., Pickles, A., Rutter, M., 2000. The autism K.L., Carter, C.S., 2012. Neuroendocrine and behavioural diagnostic observation schedule-generic: a standard measure 216 brain research 1580 (2014) 199–218

of social and communication deficits associated with the mouse lines relevant to autism. Genes Brain. Behav. 8, spectrum of autism. J. Autism Dev. Disord. 30, 205–223. 129–142. Lowery, M.C., Morris, C.A., Ewart, A., Brothman, L.J., Zhu, X.L., Munesue, T., Yokoyama, S., Nakamura, K., Anitha, A., Yamada, K., Leonard, C.O., Carey, J.C., Keating, M., Brothman, A.R., 1995. Hayashi, K., Asaka, T., Liu, H.X., Jin, D., Koizumi, K., Islam, M. Strong correlation of elastin deletions, detected by FISH, with S., Huang, J.J., Ma, W.J., Kim, U.H., Kim, S.J., Park, K., Kim, D., Williams syndrome: evaluation of 235 patients. Am. J. Hum. Kikuchi, M., Ono, Y., Nakatani, H., Suda, S., Miyachi, T., Hirai, Genet. 57, 49–53. H., Salmina, A., Pichugina, Y.A., Soumarokov, A.A., Takei, N., Macdonald, K., Macdonald, T.M., 2010. The peptide that binds: a Mori, N., Tsujii, M., Sugiyama, T., Yagi, K., Yamagishi, M., systematic review of oxytocin and its prosocial effects in Sasaki, T., Yamasue, H., Kato, N., Hashimoto, R., Taniike, M., humans. Harv. Rev. Psychiatry 18, 1–21. Hayashi, Y., Hamada, J., Suzuki, S., Ooi, A., Noda, M., Macdonald, K., Feifel, D., 2013. Helping oxytocin deliver: Kamiyama, Y., Kido, M.A., Lopatina, O., Hashii, M., Amina, S., considerations in the development of oxytocin-based Malavasi, F., Huang, E.J., Zhang, J., Shimizu, N., Yoshikawa, T., therapeutics for brain disorders. Front. Neurosci. 7, 35. Matsushima, A., Minabe, Y., Higashida, H., 2010. Two genetic Maes, B., Fryns, J.P., Van Walleghem, M., Van den Berghe, H., 1994. variants of CD38 in subjects with autism spectrum disorder Cognitive functioning and information processing of adult and controls. Neurosci. Res. 67, 181–191. mentally retarded men with fragile-X syndrome. Am. J. Med. Muscatelli, F., Abrous, D.N., Massacrier, A., Boccaccio, I., Le Moal, Genet. 50, 190–200. M., Cau, P., Cremer, H., 2000. Disruption of the mouse Necdin Martin, A., State, M., Anderson, G.M., Kaye, W.M., Hanchett, J.M., gene results in hypothalamic and behavioral alterations McConaha, C.W., North, W.G., Leckman, J.F., 1998. reminiscent of the human Prader–Willi syndrome. Hum. Mol. Cerebrospinal fluid levels of oxytocin in Prade–Willi Genet. 9, 3101–3110. syndrome: a preliminary report. Biol. Psychiatry 44, 1349–1352. Neumann, I.D., Landgraf, R., 2012. Balance of brain oxytocin and McCracken, J.T., McGough, J., Shah, B., Cronin, P., Hong, D., Aman, vasopressin: implications for anxiety, depression, and social M.G., Arnold, L.E., Lindsay, R., Nash, P., Hollway, J., McDougle, behaviors. Trends Neurosci. 35, 649–659. C.J., Posey, D., Swiezy, N., Kohn, A., Scahill, L., Martin, A., Nielsen, D.M., Derber, W.J., McClellan, D.A., Crnic, L.S., 2002. Koenig, K., Volkmar, F., Carroll, D., Lancor, A., Tierney, E., Alterations in the auditory startle response in Fmr1 targeted Ghuman, J., Gonzalez, N.M., Grados, M., Vitiello, B., Ritz, L., mutant mouse models of fragile X syndrome. Brain Res. 927, Davies, M., Robinson, J., McMahon, D., 2002. Risperidone in 8–17. children with autism and serious behavioral problems. N. O’Donnell, W.T., Warren, S.T., 2002. A decade of molecular studies Engl. J. Med. 347, 314–321. of Fragile X syndrome. Annu. Rev. Neurosci. 25, 315–338. McCullough, M.E., Churchland, P.S., Mendez, A.J., 2013. Problems Owley, T., Brune, C.W., Salt, J., Walton, L., Guter, S., Ayuyao, N., with measuring peripheral oxytocin: can the data on oxytocin Gibbons, R.D., Leventhal, B.L., Cook, E.H., 2010. A and human behavior be trusted?. Neurosci. Biobehav. Rev. 37, pharmacogenetic study of escitalopram in autism spectrum 1485–1492. disorders. Autism Res. 3, 1–7. McLennan, Y., Polussa, J., Tassone, F., Hagerman, R., 2011. Fragile Paradee, W., Melikian, H.E., Rasmussen, D.L., Kenneson, A., Conn, X syndrome. Curr. Genomics 12, 216–224. P.J., Warren, S.T., 1999. Fragile X mouse: strain effects of Mervis, C.B., Morris, C.A., Klein-Tasman, B.P., Bertrand, J., Kwitny, knockout phenotype and evidence suggesting deficient S., Appelbaum, L.G., Rice, C.E., 2003. Attentional amygdala function. Neuroscience 94, 185–192. characteristics of infants and toddlers with Williams Paxinos, G., Franklin, K.B.J., 2004. In: The Mouse Brain in syndrome during triadic interactions. Dev. Neuropsychol. 23, Stereotaxic Coordinates. Elsevier Academic Press, 243–268. Amsterdam; Boston. Miller, M., Bales, K.L., Taylor, S.L., Yoon, J., Hostetler, C.M., Carter, Peier, A.M., McIlwain, K.L., Kenneson, A., Warren, S.T., Paylor, R., C.S., Solomon, M., 2013. Oxytocin and vasopressin in children Nelson, D.L., 2000. Over)correction of FMR1 deficiency with and adolescents with autism spectrum disorders: sex differences and associations with symptoms. Autism Res. 6, YAC transgenics: behavioral and physical features. Hum. Mol. 91–102. Genet. 9, 1145–1159. Mineur, Y.S., Sluyter, F., de Wit, S., Oostra, B.A., Crusio, W.E., 2002. Petrovic, P., Kalisch, R., Singer, T., Dolan, R.J., 2008. Oxytocin Behavioral and neuroanatomical characterization of the Fmr1 attenuates affective evaluations of conditioned faces and knockout mouse. Hippocampus 12, 39–46. amygdala activity. J. Neurosci. 28, 6607–6615. Mineur, Y.S., Huynh, L.X., Crusio, W.E., 2006. Social behavior Pietropaolo, S., Guilleminot, A., Martin, B., D’Amato, F.R., Crusio, deficits in the Fmr1 mutant mouse. Behav. Brain Res. 168, W.E., 2011. Genetic-background modulation of core and 172–175. variable autistic-like symptoms in Fmr1 knock-out mice. PLoS Modahl, C., Green, L., Fein, D., Morris, M., Waterhouse, L., One 6, e17073. Feinstein, C., Levin, H., 1998. Plasma oxytocin levels in autistic Plesa-Skwerer, D., Faja, S., Schofield, C., Verbalis, A., Tager- children. Biol. Psychiatry 43, 270–277. Flusberg, H., 2006. Perceiving facial and vocal expressions of Morris, C.A., 1993. Williams Syndrome. In: Pagon, R.A., Adam, M. emotion in individuals with Williams syndrome. Am. J. Ment. P., Bird, T.D., Dolan, C.R., Fong, C.T., Stephens, K., (Eds). Retard. 111, 15–26. GeneReviews, Seattle (WA). Pretorius, P.H., King, M.A., Pan, T.S., de Vries, D.J., Glick, S.J., Byrne, Morton, J.E., Bundey, S., Webb, T.P., MacDonald, F., Rindl, P.M., C.L., 1998. Reducing the influence of the partial volume effect Bullock, S., 1997. Fragile X syndrome is less common than on SPECT activity quantitation with 3D modelling of spatial previously estimated. J. Med. Genet. 34, 1–5. resolution in iterative reconstruction. Phys. Med. Biol. 43, Moy, S.S., Nadler, J.J., Young, N.B., Perez, A., Holloway, L.P., 407–420. Barbaro, R.P., Barbaro, J.R., Wilson, L.M., Threadgill, D.W., Proulx, E., Young, E.J., Osborne, L.R., Lambe, E.K., 2010. Enhanced Lauder, J.M., Magnuson, T.R., Crawley, J.N., 2007. Mouse prefrontal serotonin 5-HT(1A) currents in a mouse model of behavioral tasks relevant to autism: phenotypes of 10 inbred Williams–Beuren syndrome with low innate anxiety. J. strains. Behav. Brain Res. 176, 4–20. Neurodev. Disord. 2, 99–108. Moy, S.S., Nadler, J.J., Young, N.B., Nonneman, R.J., Grossman, A. Reiss, A.L., Abrams, M.T., Greenlaw, R., Freund, L., Denckla, M.B., W., Murphy, D.L., D’Ercole, A.J., Crawley, J.N., Magnuson, T.R., 1995. Neurodevelopmental effects of the FMR-1 full mutation Lauder, J.M., 2009. Social approach in genetically engineered in humans. Nat. Med. 1, 159–167. brain research 1580 (2014) 199–218 217

Riby, D., Hancock, P.J., 2009. Looking at movies and cartoons: eye- Schneiderman, I., Zagoory-Sharon, O., Leckman, J.F., Feldman, R., tracking evidence from Williams syndrome and autism. J. 2012. Oxytocin during the initial stages of romantic Intellect. Disabil. Res. 53, 169–181. attachment: relations to couples’ interactive reciprocity. Rimmele, U., Hediger, K., Heinrichs, M., Klaver, P., 2009. Oxytocin Psychoneuroendocrinology 37, 1277–1285. makes a face in memory familiar. J. Neurosci. 29, 38–42. Schradin, C., Kenkel, W., Krackow, S., Carter, C.S., 2013. Staying Rodgers, J., Riby, D.M., Janes, E., Connolly, B., McConachie, H., put or leaving home: endocrine, neuroendocrine and 2012. Anxiety and repetitive behaviours in autism spectrum behavioral consequences in male African striped mice. Horm. disorders and williams syndrome: a cross-syndrome Behav. 63, 136–143. comparison. J. Autism Dev. Disord. 42, 175–180. Schubert, C., 2009. The genomic basis of the Williams-Beuren Rudelli, R.D., Brown, W.T., Wisniewski, K., Jenkins, E.C., Laure- syndrome. Cell Mol. Life Sci. 66, 1178–1197. Kamionowska, M., Connell, F., Wisniewski, H.M., 1985. Adult Schultz, R.T., 2005. Developmental deficits in social perception in fragile X syndrome. Clinico-neuropathologic findings. Acta autism: the role of the amygdala and fusiform face area. Int. J. Neuropathol. 67, 289–295. Dev. Neurosci. 23, 125–141. Ryan, B.C., Young, N.B., Crawley, J.N., Bodfish, J.W., Moy, S.S., 2010. Seltzer, L.J., Ziegler, T.E., Pollak, S.D., 2010. Social vocalizations Social deficits, stereotypy and early emergence of repetitive can release oxytocin in humans. Proc. Biol. Sci. 277, behavior in the C58/J inbred mouse strain. Behav. Brain Res. 2661–2666. 208, 178–188. Shamay-Tsoory, S.G., Fischer, M., Dvash, J., Harari, H., Perach- Saitoh, S., Buiting, K., Cassidy, S.B., Conroy, J.M., Driscoll, D.J., Bloom, N., Levkovitz, Y., 2009. Intranasal administration of Gabriel, J.M., Gillessen-Kaesbach, G., Glenn, C.C., Greenswag, oxytocin increases envy and schadenfreude (gloating). Biol. L.R., Horsthemke, B., Kondo, I., Kuwajima, K., Niikawa, N., Psychiatry 66, 864–870. Rogan, P.K., Schwartz, S., Seip, J., Williams, C.A., Nicholls, R.D., Silverman, J.L., Yang, M., Turner, S.M., Katz, A.M., Bell, D.B., 1997. Clinical spectrum and molecular diagnosis of Angelman Koenig, J.I., Crawley, J.N., 2010. Low stress reactivity and and Prader–Willi syndrome patients with an imprinting neuroendocrine factors in the BTBR T tf/J mouse model of þ mutation. Am. J. Med. Genet. 68, 195–206. autism. Neuroscience 171, 1197–1208. Sakurai, T., Dorr, N.P., Takahashi, N., McInnes, L.A., Elder, G.A., Silverman, J.L., Smith, D.G., Rizzo, S.J., Karras, M.N., Turner, S.M., Buxbaum, J.D., 2011. Haploinsufficiency of Gtf2i, a gene Tolu, S.S., Bryce, D.K., Smith, D.L., Fonseca, K., Ring, R.H., deleted in Williams Syndrome, leads to increases in social Crawley, J.N., 2012. Negative allosteric modulation of the interactions. Autism Res. 4, 28–39. mGluR5 receptor reduces repetitive behaviors and rescues Sala, M., Braida, D., Lentini, D., Busnelli, M., Bulgheroni, E., social deficits in mouse models of autism. Sci. Transl. Med. 4, Capurro, V., Finardi, A., Donzelli, A., Pattini, L., Rubino, T., 131ra51. Parolaro, D., Nishimori, K., Parenti, M., Chini, B., 2011. Spencer, C.M., Alekseyenko, O., Serysheva, E., Yuva-Paylor, L.A., Pharmacologic rescue of impaired cognitive flexibility, social Paylor, R., 2005. Altered anxiety-related and social behaviors deficits, increased aggression, and seizure susceptibility in in the Fmr1 knockout mouse model of fragile X syndrome. oxytocin receptor null mice: a neurobehavioral model of Genes, Brain Behav. 4, 420–430. autism. Biol. Psychiatry 69, 875–882. Swaab, D.F., Purba, J.S., Hofman, M.A., 1995. Alterations in the Sanders, S.J., Ercan-Sencicek, A.G., Hus, V., Luo, R., Murtha, M.T., hypothalamic paraventricular nucleus and its oxytocin Moreno-De-Luca, D., Chu, S.H., Moreau, M.P., Gupta, A.R., neurons (putative satiety cells) in Prader–Willi syndrome: a Thomson, S.A., Mason, C.E., Bilguvar, K., Celestino-Soper, P.B., study of five cases. J. Clin. Endocrinol. Metab. 80, 573–579. Choi, M., Crawford, E.L., Davis, L., Wright, N.R., Dhodapkar, R. Szeto, A., McCabe, P.M., Nation, D.A., Tabak, B.A., Rossetti, M.A., M., DiCola, M., DiLullo, N.M., Fernandez, T.V., Fielding-Singh, McCullough, M.E., Schneiderman, N., Mendez, A.J., 2011. V., Fishman, D.O., Frahm, S., Garagaloyan, R., Goh, G.S., Evaluation of enzyme immunoassay and radioimmunoassay Kammela, S., Klei, L., Lowe, J.K., Lund, S.C., McGrew, A.D., methods for the measurement of plasma oxytocin. Meyer, K.A., Moffat, W.J., Murdoch, J.D., O’Roak, B.J., Ober, G.T., Psychosom. Med. 73, 393–400. Pottenger, R.S., Raubeson, M.J., Song, Y., Wang, Q., Yaspan, B. Takayanagi, Y., Yoshida, M., Bielsky, I.F., Ross, H.E., Kawamata, M., L., Yu, T.W., Yurkiewicz, I.R., Beaudet, A.L., Cantor, R.M., Onaka, T., Yanagisawa, T., Kimura, T., Matzuk, M.M., Young, L. Curland, M., Grice, D.E., Gunel, M., Lifton, R.P., Mane, S.M., J., Nishimori, K., 2005. Pervasive social deficits, but normal Martin, D.M., Shaw, C.A., Sheldon, M., Tischfield, J.A., Walsh, parturition, in oxytocin receptor-deficient mice. Proc. Natl. C.A., Morrow, E.M., Ledbetter, D.H., Fombonne, E., Lord, C., Acad. Sci. USA 102, 16096–16101 (Epub 2005 Oct 25). Martin, C.L., Brooks, A.I., Sutcliffe, J.S., Cook Jr., E.H., Tassone, F., Hagerman, R.J., Chamberlain, W.D., Hagerman, P.J., Geschwind, D., Roeder, K., Devlin, B., State, M.W., 2011. 2000a. Transcription of the FMR1 gene in individuals with Multiple recurrent de novo CNVs, including duplications of fragile X syndrome. Am. J. Med. Genet. 97, 195–203. the 7q11.23 Williams syndrome region, are strongly associated Tassone, F., Hagerman, R.J., Taylor, A.K., Gane, L.W., Godfrey, T.E., with autism. Neuron 70, 863–885. Hagerman, P.J., 2000b. Elevated levels of FMR1 mRNA in carrier Sauer, C., Montag, C., Worner, C., Kirsch, P., Reuter, M., 2012. males: a new mechanism of involvement in the fragile-X Effects of a common variant in the CD38 gene on social syndrome. Am. J. Hum. Genet. 66, 6–15. processing in an oxytocin challenge study: possible links to Tauber, M., Mantoulan, C., Copet, P., Jauregui, J., Demeer, G., autism. Neuropsychopharmacology 37, 1474–1482. Diene, G., Roge, B., Laurier, V., Ehlinger, V., Arnaud, C., Sawchenko, P.E., Swanson, L.W., 1985. Localization, Molinas, C., Thuilleaux, D., 2011. Oxytocin may be useful to colocalization, and plasticity of corticotropin-releasing factor increase trust in others and decrease disruptive behaviours in immunoreactivity in rat brain. Fed. Proc. 44, 221–227. patients with Prader–Willi syndrome: a randomised placebo- Schain, R.J., Freedman, D.X., 1961. Studies on 5-hydroxyindole controlled trial in 24 patients. Orphanet J. Rare Dis. 6, 47. metabolism in autistic and other mentally retarded children. J. Teng, B.L., Nonneman, R.J., Agster, K.L., Nikolova, V.D., Davis, T.T., Pediatr. 58, 315–320. Riddick, N.V., Baker, L.K., Pedersen, C.A., Jarstfer, M.B., Moy, S. Schapiro, M.B., Murphy, D.G., Hagerman, R.J., Azari, N.P., S., 2013. Prosocial effects of oxytocin in two mouse models of Alexander, G.E., Miezejeski, C.M., Hinton, V.J., Horwitz, B., autism spectrum disorders. Neuropharmacology 72, 187–196. Haxby, J.V., Kumar, A., et al., 1995. Adult fragile X syndrome: The Dutch-Belgian Fragile-X Consortium, 1994. Fmr1 knockout neuropsychology, brain anatomy, and metabolism. Am. J. Med. mice: a model to study fragile X mental retardation. The Genet. 60, 480–493. Dutch-Belgian Fragile X Consortium. Cell 78, 23–33. 218 brain research 1580 (2014) 199–218

Tsai, T.F., Jiang, Y.H., Bressler, J., Armstrong, D., Beaudet, A.L., Witt, D.M., Insel, T.R., 1992. Central oxytocin antagonism 1999. Paternal deletion from Snrpn to Ube3a in the mouse decreases female reproductive behavior. Ann. N. Y. Acad. Sci. causes hypotonia, growth retardation and partial lethality and 652, 445–447. provides evidence for a gene contributing to Prader–Willi Witt, D.M., Winslow, J.T., Insel, T.R., 1992. Enhanced social syndrome. Hum. Mol. Genet. 8, 1357–1364. interactions in rats following chronic, centrally infused Turner, G., Webb, T., Wake, S., Robinson, H., 1996. Prevalence of oxytocin. Pharmacol. Biochem. Behav. 43, 855–861. fragile X syndrome. Am. J. Med. Genet. 64, 196–197. Wohrle, D., Kotzot, D., Hirst, M.C., Manca, A., Korn, B., Schmidt, Vorstman, J.A., Staal, W.G., van Daalen, E., van Engeland, H., A., Barbi, G., Rott, H.D., Poustka, A., Davies, K.E., et al., 1992. A Hochstenbach, P.F., Franke, L., 2006. Identification of novel microdeletion of less than 250 kb, including the proximal part autism candidate regions through analysis of reported of the FMR-I gene and the Fragile-X site, in a male with the cytogenetic abnormalities associated with autism. Mol. clinical phenotype of fragile-X syndrome. Am. J. Hum. Genet. Psychiatry 11 (1), 18–28. 51, 299–306. Wang, L.W., Berry-Kravis, E., Hagerman, R.J., 2010. Fragile X: Wu, S., Jia, M., Ruan, Y., Liu, J., Guo, Y., Shuang, M., Gong, X., leading the way for targeted treatments in autism. Zhang, Y., Yang, X., Zhang, D., 2005. Positive association of the Neurotherapeutics 7, 264–274. oxytocin receptor gene (OXTR) with autism in the Chinese Waterhouse, L., Fein, D., Modahl, C., 1996. Neurofunctional Han population. Biol. Psychiatry 58, 74–77. mechanisms in autism. Psychol. Rev. 103, 457–489. Xu, X.J., Shou, X.J., Li, J., Jia, M.X., Zhang, J.S., Guo, Y., Wei, Q.Y., Webb, S., 2010. Drugmakers dance with autism. Nat. Biotechnol. Zhang, X.T., Han, S.P., Zhang, R., Han, J.S., 2013. Mothers 28, 772–774. of autistic children: lower plasma levels of oxytocin and Weiler, I.J., Irwin, S.A., Klintsova, A.Y., Spencer, C.M., Brazelton, A. arg-vasopressin and a higher level of testosterone. PLoS One 8, D., Miyashiro, K., Comery, T.A., Patel, B., Eberwine, J., e74849. Greenough, W.T., 1997. Fragile X mental retardation protein is Yamamoto, Y., Cushing, B.S., Kramer, K.M., Epperson, P.D., translated near synapses in response to neurotransmitter Hoffman, G.E., Carter, C.S., 2004. Neonatal manipulations of activation. Proc. Natl. Acad. Sci. USA 94, 5395–5400. oxytocin alter expression of oxytocin and vasopressin Weiler, I.J., Spangler, C.C., Klintsova, A.Y., Grossman, A.W., Kim, S. immunoreactive cells in the paraventricular nucleus of the H., Bertaina-Anglade, V., Khaliq, H., de Vries, F.E., Lambers, F. hypothalamus in a gender-specific manner. Neuroscience 125, A., Hatia, F., Base, C.K., Greenough, W.T., 2004. Fragile X 947–955. mental retardation protein is necessary for neurotransmitter- Yan, Q.J., Asafo-Adjei, P.K., Arnold, H.M., Brown, R.E., Bauchwitz, activated protein translation at synapses. Proc. Natl. Acad. Sci. R.P., 2004. A phenotypic and molecular characterization of the USA 101, 17504–17509. fmr1-tm1Cgr fragile X mouse. Genes Brain Behav. 3, 337–359. Whitman, B.Y., Accardo, P., 1987. Emotional symptoms in Prader– Willi syndrome adolescents. Am. J. Med. Genet. 28, 897–905. Yoshida, M., Takayanagi, Y., Inoue, K., Kimura, T., Young, L.J., Whittington, J., Holland, T., 2011. Recognition of emotion in facial Onaka, T., Nishimori, K., 2009. Evidence that oxytocin exerts expression by people with Prader–Willi syndrome. J. Intellect. anxiolytic effects via oxytocin receptor expressed in Disabil. Res. 55, 75–84. serotonergic neurons in mice. J. Neurosci. 29, 2259–2271. Williams, J.C., Barratt-Boyes, B.G., Lowe, J.B., 1961. Supravalvular Young, L.J., Flanagan-Cato, L.M., 2012. Editorial comment: aortic stenosis. Circulation 24, 1311–1318. oxytocin, vasopressin and social behavior. Horm. Behav. 61, Wilson, B.M., Cox, C.L., 2007. Absence of metabotropic glutamate 227–229. receptor-mediated plasticity in the neocortex of fragile X Yrigollen, C.M., Han, S.S., Kochetkova, A., Babitz, T., Chang, J.T., mice. Proc. Natl. Acad. Sci. USA 104, 2454–2459. Volkmar, F.R., Leckman, J.F., Grigorenko, E.L., 2008. Genes Winslow, J.T., Insel, T.R., 2004. Neuroendocrine basis of social controlling affiliative behavior as candidate genes for autism. recognition. Curr. Opin. Neurobiol. 14, 248–253. Biol. Psychiatry 63, 911–916. Wismer Fries, A.B., Ziegler, T.E., Kurian, J.R., Jacoris, S., Pollak, S. Zhao, M.G., Toyoda, H., Ko, S.W., Ding, H.K., Wu, L.J., Zhuo, M., D., 2005. Early experience in humans is associated with 2005. Deficits in trace fear memory and long-term changes in neuropeptides critical for regulating social potentiation in a mouse model for fragile X syndrome. J. behavior. Proc. Natl. Acad. Sci. USA. 102, 17237–17240. Neurosci. 25, 7385–7392. Wisniewski, K.E., Segan, S.M., Miezejeski, C.M., Sersen, E.A., Zhong, S., Monakhov, M., Mok, H.P., Tong, T., Lai, P.S., Chew, S.H., Rudelli, R.D., 1991. The Fra(X) syndrome: neurological, Ebstein, R.P., 2012. U-shaped relation between plasma electrophysiological, and neuropathological abnormalities. oxytocin levels and behavior in the trust game. PLoS One 7, Am. J. Med. Genet. 38, 476–480. e51095. brain research 1580 (2014) 172–179

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Oxytocin differentially modulates compromise and competitive approach but not withdrawal to antagonists from own vs. rivaling other groups

Femke S. Ten Veldena, Matthijs Baasa, Shaul Shalvib, Mariska E. Kreta, Carsten K.W. De Dreua,n aCognitive Science Center Amsterdam, University of Amsterdam, Weesperplein 4, 1018XA Amsterdam, The Netherlands bDepartment of Psychology, Ben Gurion University of the Negev, Beer Sheva 84105, Israel article info abstract

Article history: In humans, oxytocin promotes cognitive and motivational tendencies that benefit the groups Accepted 15 September 2013 on which humans depend for their survival and prosperity. Here we examined decision Available online 20 September 2013 making in an incentivized two-player poker game with either an in-group or out-group Keywords: antagonist. Sixty nine healthy males received 24 IU oxytocin or matching placebo, and played fi Social preferences four rounds of a simpli ed poker game. On each round they received either low or high value Conflict cards to create differences in competitive strength, and then responded to a bet placed by Cooperation their (simulated) (in-group or out-group) antagonist. Under placebo, participants withdrew Human groups and competed depending on their own (low vs. high) competitive strength, regardless of their Hormones antagonist's group membership. Under oxytocin, however, participants settled more and competed less with an in-group as compared to an out-group antagonist; withdrawal was unaffected by group membership. We conclude that oxytocin sensitizes humans to the group membership of their interaction partner, rendering them relatively more benevolent and less competitive towards those seen as belonging to their own group. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2013 Elsevier B.V. All rights reserved.

1. Introduction (Kirsch et al., 2005). Oxytocin interacts with dopaminergic, reward processing circuits in the nucleus accumbens shell In their search for the neurobiological bases of social behavior, and in the ventral tegmental area (Skuse and Gallagher, 2008), scientists across the behavioral and brain sciences turned their and exerts anxiolytic effects viadirectactivationofoxytocin focus to oxytocin, an evolutionary ancient and structurally receptors expressed in serotonergic neurons of the raphe nuclei highly preserved neuropeptide (e.g., Bartz et al., 2010; Chang (Veenema et al., 2010; Yoshida et al., 2009). et al., 2012; Bos et al., 2012; Ross and Young, 2009; Striepens et al., 2012). Oxytocin is produced in the hypothalamus and 1.1. Social bond formation and maintenance released into the blood stream from axon terminals and into the brain from dendrites of hypothalamic neurons (Donaldson Oxytocin is perhaps best known for its critical role in and Young, 2008; Ludwig and Leng, 2006). Functioning as both parturition and reproduction on the one hand, and social aneurotransmitterandhormone,oxytocin'stargetsare bond formation and maintenance on the other (e.g., Carter widespread and include the hippocampus and the amygdala et al., 2008). First, male rodents engineered to lack (fore-brain)

nCorresponding author. E-mail addresses: [email protected] (F.S. Ten Velden), [email protected] (M. Baas), [email protected] (S. Shalvi), [email protected] (M.E. Kret), [email protected] (C.K.W. De Dreu).

0006-8993/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.brainres.2013.09.013 brain research 1580 (2014) 172–179 173

oxytocin receptors no longer discriminate between familiar with those classified as rivaling out-group members (De Dreu and unfamiliar females, whereas normal rodents spent more et al., 2011, 2012b; Sheng et al., 2013). Third, individuals given time investigating unfamiliar female rodents vs. female oxytocin rather than placebo conform to opinions of in-group rodents with whom they had shared a cage for several days members more than to (identical) opinions voiced by out- (Macbeth et al., 2009; also see Ferguson et al., 2000, 2002). group members (Stallen et al., 2012). Fourth, individuals given Along similar lines, participants who memorized pictures of oxytocin self-sacrifice more, and contribute to their own faces under oxytocin performed better one day later on group more than to the broader collective that includes both measures of familiarity, indicating that oxytocin makes a their own group and other groups (Israel et al., 2012). Finally, face in memory more familiar (Rimmele et al., 2009). when their own group competes with an out-group, indivi- In addition to social bond formation, oxytocin also appears duals given oxytocin prefer strong allies (De Dreu et al., 2012a; to stimulate empathic responding, which is important also to also see Kret and De Dreu, 2013) and display parochial social bond maintenance. For example, in women exposed altruism – a tendency to cooperate with the in-group and to to infant crying, intranasal oxytocin modulates activity in compete against the out-group (De Dreu et al., 2010, 2012b; the inferior frontal gyrus (Riem et al., 2011), fathers given also see Choi and Bowles, 2007; Israel et al., 2012). oxytocin rather than placebo stimulate their toddler's explora- Taken together, it thus stands to reason that oxytocin does tion more and show less hostility (Naber et al., 2010), and in not promote indiscriminate pro-social tendencies. Instead, it males exposed to biological motion (a point-light figure repre- appears that oxytocin promotes cognitive, motivational, and senting a walking human), intranasal oxytocin modulates behavioral tendencies that are beneficial to the groups within neural circuitries involved in affective perspective taking which humans operate and upon which they depend for (Keri and Benedek, 2009; Perry et al., 2010). Other studies survival and prosperity (De Dreu, 2012; van Ijzendoorn and showed that participants given oxytocin rather than placebo Bakermans-Kranenburg, 2012). Such tendencies include have increased sensitivity to other's fear (Fischer-Shofty et al., in-group love and, if necessary for in-group protection, out- 2010), empathize more with people depicted in emotionally group hate as well. charged situations (Hurlemann et al., 2010; but see Singer et al., 2008), and more accurately infer emotions expressed by others 1.3. Current study: decision making in competitive (Domes et al., 2007). Indeed, both humans and non-human interactions mammals show increased benevolence under oxytocin, includ- ing tendencies to benefit con-specifics (Chang et al., 2012), to The conclusion that oxytocin promotes group-serving ten- trust others (Baumgartner et al., 2008; Kosfeld et al., 2005), to dencies rests on studies examining relatively cooperative make fair offers in bargaining (Zak et al., 2007), and to benefit situations where humans faced the choice to contribute to others at a personal cost (e.g., Morhenn et al., 2008). their group or not, to trust others or not, or to make (un)fair offers. However, in addition to these more benign situations, 1.2. Indiscriminate benevolence vs. group-serving group life is marked also by conflict when, for example, tendencies individuals compete for status and scarce resources. Typi- cally, such conflicts trigger a tendency towards (i) withdrawal Whereas the heretofore reviewed work suggest that oxytocin and subordination, (ii) matching and compromise, or (iii) promotes indiscriminate benevolence and generosity (e.g., aggressive approach (De Dreu, 2010; Deutsch, 1973). Although Zak et al., 2007), a more accurate conclusion appears that individuals have an incentive to compete through aggressive oxytocin promotes group-serving tendencies (De Dreu, 2012; approach, their overarching group fares better when conflict Goodson, 2013). For example, meerkats live in clans and their is mitigated through withdrawal and compromise (De Dreu, survival and prosperity depends on successful in-clan coop- 2010). Accordingly, our conjecture that oxytocin promotes eration and coordination and defense to predators and roving group-serving tendencies implies that in competitive inter- competing clans (Drewe et al., 2009). In free-living meerkats, actions, oxytocin increases (i) costly withdrawal and/or peripheral administration of oxytocin rather than placebo settlement, and (ii) reduces aggressive approach, especially increased an array of cooperative behaviors directed at the when (iii) antagonists are part of one's in-group rather than own clan, including digging, associating with pups, and time- coming from rivaling out-groups. on-guard (Madden and Clutton-Brock, 2011). Other studies We examined this possibility in an incentivized two- documented that oxytocin is key in triggering so-called player poker game adapted from Ten Velden et al. (2012). maternal defense, which occurs when a breast-feeding mother Fig. 1 provides a schematic overview of the experimental is faced with an unfamiliar intruder and lashes out to protect procedures (see section 4 for further detail). and defend its pups (Bosch et al., 2005; Pedersen et al., 1982). Participants received oxytocin or matching placebo, and In humans, similar tendencies have been documented as were paired to a (simulated) antagonist from their own well. First, the hypothalamic release of oxytocin is promoted in-group, or from a rivaling out-group. In this simplified by displays of trust and cooperation by others, especially poker-game, participants are given chips with monetary value, familiar others like parents and intimate partners (e.g., Ditzen and handed a card from a 52-card deck. Following an initial et al., 2007; Feldman et al., 2010; Gordon et al., 2010; Holt- forced bet which starts the game, participants observe their Lunstad et al., 2008; Morhenn et al., 2008; Uvnas-Moberg, antagonist's bet, to which they may respond by withdrawing 1998; Zak et al., 2005). Second, when given oxytocin rather from the game at a personal cost (i.e., “fold,” where the bet is than placebo, humans display more positive attitudes and lost and the pot is handed to the opponent), by (ii) compromising empathize only with members of their own group and not (i.e., “call;” they match their antagonist's bet and the player 174 brain research 1580 (2014) 172–179

Fig. 1 – Time-line of the experiment; tasks taking place within individual cubicles were entirely computer-guided and participants worked alone at their own pace.

2 with the highest card wins the pot; in case of a draw, the pot F(1, 65) 156.76, p 0.001, ηp 0.707, and an unexpected com- ¼ ¼ ¼ is split between participant and antagonist), or (iii) competing petitive strength treatment effect, F(1, 65) 4.51, p 0.037, 2  ¼ ¼ (i.e., “raise,” in which case the participant increases the size of ηp 0.065. Inspection of the means showed that in both ¼ the pot). Raising is considered an aggressive approach strategy treatment conditions, low value cards induced less perceived with which more chips can be gained or lost (depending on the competitive strength than high value cards, but this effect strength of the card) if the competitor calls the bet; chips can be was weaker among participants given oxytocin (M 2.14, ¼ gained if the competitor folds. In terms of our hypotheses, SD 0.72 vs. M 3.51, SD 0.92) rather than placebo (M 1.82, ¼ ¼ ¼ ¼ we expected individuals to fold and/or call more, and to raise SD 0.78 vs. M 3.75, SD 0.73). In all, however, we concluded ¼ ¼ ¼ less when confronted with an in-group (vs. out-group) antago- that our manipulation of competitive strength was successful nist, especially when given oxytocin rather than placebo. We and as intended. manipulated Competitive Strength by providing participants twice with a high value card, and twice with a low value card 2.2. Decision making (Section 4), to explore whether competitive strength modulates predicted effects of oxytocin on decision making. Across decision rounds, participants were given twice a low, and twice a high value card. For each value, we counted how often participants decided to fold, call, or raise (each ranges 2. Results between 0 and 2). Decisions were submitted to a 2(treatment) 2(antagonist's group) 3(decision: fold vs. call vs. raise) 2    2.1. Manipulation checks (competitive strength: low vs. high) Mixed Model ANOVA with the last two factors within-subjects. This revealed, first To verify the adequacy of the manipulation of the antago- of all, a main effect for Decision, F(2, 65) 38.109, p 0.001, 2 ¼ ¼ nist's group membership, we cross-tabulated responses to ηp 0.543, and a Decision Competitive Strength interaction, ¼  2 the question whether the antagonist was a member of one's F(2, 65) 204.301, p 0.0001, ηp 0.864. Fig. 2 shows that when ¼ ¼ ¼ own group, or of the other group (1 my team, 2 other team) given low value cards, participants more often decided to fold ¼ ¼ with Antagonist's Group, Treatment, and Group Treatment (M 1.565, SD 0.606), and less often to raise (M 0.217, SD  ¼ ¼ ¼ ¼ interaction. Responses only differed as a function of Antago- 0.449), than when given high value cards (M 0.014, SD 0.12 ¼ ¼ nist Group, χ2(1, 69) 68.108, p 0.0001. Except for one mis- for fold; M 1.594, SD 0.577 for raise). Decisions to call did ¼ ¼ ¼ ¼ take, the antagonist's group membership was accurately not change as a function of competitive strength (M 0.217, ¼ identified. SD 0.449 for low value; M 0.391, SD 0.548 for high value). ¼ ¼ ¼ The competitive strength manipulation was verified after Treatment did not interact with Decision, F(2, 65) 1.014, ¼ fi 2 each decision, rst, by asking participants to indicate on a p 0.368, ηp 0.031, indicating no overall preference change ¼ ¼ slider whether their card was better ( 60) to worse ( 60) than induced by oxytocin. However, as predicted, we did observe a þ À their antagonist's card. Ratings were averaged across low significant Treatment Decision Antagonist's Group inter-  2 value and high value cards, respectively, and submitted to a 2 action, F(2, 65) 4.045, p 0.022, ηp 0.112. Follow-up analyses ¼ ¼ ¼ (treatment) 2(antagonist's group) 2(high vs. low competi- using the overall error term and associated degrees of free-   tive strength) Mixed-Model ANOVA with competitive strength dom (Tatsuoka, 1988) indicated that whereas Decision and as within-subjects factor. Results showed only a main effect Antagonist's Group did not interact under placebo, F(2, 65) 2 2 ¼ for competitive strength, F(1, 65) 373.13, p 0.0001, ηp 0.852, 0.978, p 0.381, ηp 0.050, they did under oxytocin, F(2, 65) ¼ ¼ ¼ ¼ 2 ¼ ¼ indicating that participants perceived themselves to be 3.848, p 0.026, ηp 0.214. Fig. 3 (left panel) shows that when ¼ ¼ weaker than their antagonist when given a low value card participants received placebo, their decision to fold, call, or raise (M 31.173, SD 13.488, one-sample t-test, t(69) -20.605, did not depend on their antagonist's group membership. Fig. 3 ¼À ¼ ¼ p 0.001), and stronger than their antagonist when given a (right panel) shows that when participants received oxytocin, ¼ high value card (M 18.185, SD 17.466, one-sample t-test, however, they were more inclined to call when dealing with an ¼þ ¼ t(69) 8.194, p 0.001). Second, we asked participants whether in-group rather than out-group antagonist, F(1, 66) 7.447, ¼ ¼ 2 ¼ they felt they could win this round (1 not at all; to 5 very p 0.008, ηp 0.211, and less inclined to raise when dealing with ¼ ¼ ¼ ¼ much). Ratings were averaged for low value and for high an in-group rather than out-group antagonist, F(1, 66) 4.714, 2 ¼ value cards, and submitted to a 2(treatment) 2(antagonist's p 0.034, ηp 0.114. These decision patterns did not differ for  ¼ ¼ group) 2(high vs. low competitive strength) Mixed-Model low vs. high value cards, in that the Treatment Decision    ANOVA with competitive strength within-subjects. Results Antagonist's Group Competitive Strength interaction was not  revealed the expected main effect for competitive strength, significant, F(2, 64) 2.079, p 0.14. ¼ ¼ brain research 1580 (2014) 172–179 175

Fig. 3 – Decision Making as a function of Treatment and Antagonist's Group Membership range 0–4; displayed Means 7SE; Connectors indicate significant differences between means, with * po0.10 and ** po0.05 (N 69, two-tailed tests). ¼ Under placebo, antagonist's group membership has no effect Fig. 2 – Decision Making depends on competitive strength on withdrawal, compromise, or aggressive approach. Under such that individuals withdraw more and aggressively oxytocin, individuals compromise more and aggressively approach less under low competitive strength yet withdraw approach less in-group compared to out-group antagonists. less and aggressively approach more under high competitive strength (range 0–2, displayed Means 7SE). member, and do not show an increased preference for settlement when protagonists are from a rivaling out-group. Another way of looking at this complex interaction among Second, and relatedly, intranasal oxytocin promotes coopera- treatment, decision, and antagonist's group membership is tive conflict resolution in within but not between group by comparing oxytocin vs. placebo within each type of competitions. This conclusion follows from the observation decision and for each antagonist. Treatment had no effect that, compared to placebo, oxytocin increased a preference for costly withdrawal, regardless of whether the antagonist for settlement with in-group protagonists, but not with out- was in-group, F(1, 66) 2.382, p 0.127 or out-group, F(1, 66) group protagonists. Third, and finally, in competitive inter- ¼ ¼ ¼ 0.088, p 0.767. However, oxytocin compared to placebo led to actions, intranasal oxytocin does not influence preferences ¼ more compromise when dealing with an in-group antagonist, for withdrawal. This conclusion follows from the observation 2 F(1, 66) 8.837, p 0.004, ηp 0.239, and not when dealing with that regardless of competitive strength, and regardless of the ¼ ¼ ¼ 2 fl an out-group antagonist, F(1, 66) 1.037, p 0.312, ηp 0.021. protagonist's group membership, oxytocin exerted no in u- ¼ ¼ ¼ Finally, oxytocin compared to placebo reduced competitive ence whatsoever on the tendency to “fold” in the competitive approach towards an in-group, F(1, 66) 2.808, p 0.099, poker-game studied here. ¼ ¼ 2 fi ηp 0.121 (marginal), and non-signi cantly increased compe- Results corroborate our conjecture that oxytocin plays a ¼ titive approach towards an out-group, F(1, 66) 0.606, p pivotal role in creating, maintaining, and promoting humans' 2 ¼ ¼ 0.441, ηp 0.012. This corroborates earlier work showing that bonds within the groups upon which they depend. Consistent ¼ oxytocin compared to placebo increases benevolence towards with earlier studies on in-group favoritism (De Dreu et al., in-group members and not, or to a much lesser extent, motivates 2011; Sheng et al., 2013), social conformity (Stallen et al., competitiveness towards out-groups (De Dreu et al., 2010, 2011). 2012), and parochial altruism (De Dreu et al., 2010, 2012b), here we observed that oxytocin increases preferences for cooperative settlement and reduced preferences for compe- 3. Conclusions and discussion titive approach in interactions with in-group members, but not in interactions with out-group members. Results permit three conclusions. First, intranasal oxytocin The key finding that effects of oxytocin on competitive does not increase indiscriminate benevolence in humans. decision making are moderated by the antagonist's group This conclusion follows from the observation that in compe- membership subscribes to the general conclusion that, in titive interactions, humans given oxytocin reduce competi- humans, effects of oxytocin depend on both individual tive approach only when their protagonist is an in-group differences and personality characteristics, and on contextual 176 brain research 1580 (2014) 172–179

cues and situational constraints (Bartz et al., 2011). Scheele behavior, depending on the social context (Bartz et al., 2011; et al. (2012), for example, observed markedly different effects Shamay‐Tsoory et al., 2009). In the current competitive of oxytocin on interpersonal distance when participants context, this “social salience hypothesis” would imply that were romantically engaged or not on the one hand, and oxytocin may increase competitive approach when antago- the interaction target's sex. Other work observed effects of nists belong to a rivaling out-group, yet when they belong oxytocin to depend on early childhood experiences (Van to the in-group interaction oxytocin increases cooperation IJzendoorn and Bakermans-Kranenburg, in press). Similar to and reduce competitive approach. However, whereas we the current findings, it thus appears that effects of oxytocin in indeed observed that oxytocin increases in-group coopera- humans are context-dependent, and that one key context tion (increased compromise; reduced competitive involves (features of) the group setting within which the approach), it did not increase out-group competition individual operates. (reduced compromise; increased competitive approach). Our study involved healthy males engaging in a simple The final perspective on oxytocin assumes its effects on competition with a simulated protagonist. As such, current social cognition and behavior emerge because oxytocin up- conclusions are limited, potentially, to males and may not regulates social approach towards positive cues, and inhibits generalize to female participants. There is some evidence that withdrawal from negative cues (Kemp and Guastella, 2011). oxytocin up-regulates motivational attention to competition in Consistent with this is a recent study of rhesus macaques, males but not in females (Fischer-Shofty et al., in press). showing that oxytocin increases attention to faces and eyes Future research should examine current hypotheses for both and, importantly, reduces social vigilance for unfamiliar, males and females. Second, competitive interaction was stu- dominant, and emotional faces (Ebitz et al., 2013; also see died with a highly simplified poker game, in which partici- Parr et al., in press). Along similar lines, intranasal oxytocin pants made decisions without receiving feedback on the in humans facilitates recognition of and (empathic) outcomes of these decisions. Accordingly, the current study responses to positive social cues, yet does not alter recogni- contains no information about possible adaptation to the tion of, and responses to aversive social stimuli. From this, competitor's strategy, and about whether oxytocin increases Striepens et al. (2012, p. 18147) concluded that oxytocin flexible adaptation or not. There is some evidence that prepares for “approach and protective behavior, but with oxytocin, compared to placebo, promotes divergent thinking heightened caution” (p. 18147), a conclusion that fits well and flexible processing (De Dreu et al., in press), and it may the “tend-and-defend” response triggered by oxytocin in thus be possible that humans given oxytocin more flexibly intergroup competition (De Dreu et al., 2010; De Dreu, adapt to the outcomes of a competitive interaction, their 2012b). Current results also subscribe to this hypothesis. In opponent's strategy, and their combination. Herein lies competitive interactions among humans, oxytocin does not another avenue for future research into the role of oxytocin alter preferences for withdrawal, and increases pro-social in social behavior. Study limitations aside, our results inform preferences but with caution; only when dealing with in- an emerging debate about the fundamental mechanisms that group members and not with competitors from a rivaling out- account for the plethora of effects oxytocin seems to exert on group does oxytocin promote pro-social approach. social perception, social motivation, and social behavior. Three or more or less related accounts have been proposed. The first account rests on the well-documented anxiolytic effects of 4. Experimental procedures oxytocin. Oxytocin interacts with the hypothalamic-pituitary- adrenal axis to attenuate stress responses, and this has a 4.1. Participant recruitment pervasive influence throughout both the body and the brain (Neumann, 2008). Specifically, oxytocin reduces cortisol levels The study was approved by the Ethics Committee of the after exposure to stressors (Heinrichs et al., 2003), inhibits University of Amsterdam, and all participants provided cardiovascular stress responses (Uvnas-Moberg, 1998), reduces informed consent prior to participation. Male participants the activation of the amygdala and attenuates its coupling to were recruited via an on-line recruiting system and offered a brainstem centers responsible for autonomic and behavioral monetary reward of €10 (approx. 13 USD) for participating in a components of fear (Kirsch et al., 2005; Petrovic et al., 2008). study on the effects of medication on test scores and This, in turn, has been argued to allow the individual to decision-making. They filled out an on-line medical screening consider alternatives to fight-or-flight – the typical autonomic – exclusion criteria were significant medical or psychiatric response to (social) stressors – and permits pro-social illness, medication, smoking more than five cigarettes per approach (Lim and Young, 2006; Heinrichs et al., 2009; Taylor day, and drug or alcohol abuse. Seventy-three participants et al., 2000). This perspective fits the current finding that were retained and instructed to refrain from smoking or oxytocin increases a preference for settlement rather than drinking (except water) for 2 h before the experiment. A total aggressive approach with in-group competitors. However, it of sixty nine participants were included in the final sample has difficulty accounting for the observation that oxytocin did and analyses (two participants were excluded because they not reduce aggressive approach towards out-group competi- indicated themselves as very experienced poker players, and tors, a finding that emerged in other work as well (De Dreu two other participants were excluded because of technical et al., 2010, 2011; Sheng et al., 2013; also see Declerck et al., failures and missing data). There were no differences 2010; Mikolajczak et al., 2010). A second possibility is that between the two treatment groups in medical screening oxytocin increases attention to social cues and therefore it has responses. Participants averaged 21.51 (SD 3.078) years of ¼ widely varying effects on ‘downstream’ cognition and age, and age did not differ across experimental conditions, all brain research 1580 (2014) 172–179 177

F(1, 65)o0.52, all ps40.49. On average, participants indicated assigned to either Team “Triangle” or Team “Circle” on the they had moderate experience with poker (M 2.36; with basis of the order in which they signed-up for the experi- ¼ 5 very much; SD 0.94), and experience did not differ across ment, and engaged in decision making in a between-group ¼ ¼ experimental conditions, all F(1, 65)o2.10, all p40.15. competition games (De Dreu et al., 2012b). This took approxi- mately 15 min. Thereafter, they performed a short filler task 4.2. Substance administration asking about their current mood. They indicated for 8 positive and 10 negative mood states how they felt (e.g., concentrated; Participants were randomly assigned to the oxytocin or placebo 1 not at all, to 5 very much). We created a positive affect ¼ ¼ group (double-blind, placebo-controlled study design). Partici- index (8 items, α 0.67) and a negative affect index (10 items, ¼ pants self-administered a single intranasal dose of 24 IU α 0.82) and found no differences across the four conditions ¼ oxytocin (Syntocinon-Spray, Novartis; 3 puffs per nostril, each of our experimental design, all F(1, 65)o0.48, ps40.49. On with 4 IU oxytocin) or placebo 30 min before the start of the average, participants felt more positive (M 2.96, SD 0.58) ¼ ¼ experimental tasks (De Dreu et al., 2010; Kosfeld et al., 2005). than negative (M 1.33, SD 0.43). ¼ ¼ To avoid any subjective effects (for example, olfactory effects) Following the short filler task, participants were intro- other than those caused by oxytocin, the placebo contained all duced to the two-person simplified online poker game with the active ingredients except for the neuropeptide. The placebo another participant who would remain anonymous. Partici- was manufactured by Stichting Apothekers Haarlemse Zieken- pants played four rounds of poker. In each round, they huizen (SAHZ) in coordination with the pharmacy at the started with 25 chips and were reminded that the earned Amsterdam Medical Centre (AMC), adhering to the guidelines chips would be converted 1:1 into lottery tickets, which would on Good Manufacturing Practice (GMP) and Good Clinical go into a raffle for one of four 25€ Euro prizes. Before each Practice (GCP). The placebo was produced using the exact same round started, they had to place a forced bet of five chips and recipes and procedures used by Novartis Inc. to produce the read that from a standard deck of 52 playing cards, they and carrier of Syntocinon – the synthetic analog of oxytocin. their antagonist would randomly receive one card. The Placebos were delivered in the same bottles as Syntocinon. In antagonist was (unknowingly to participants) simulated and short, the only difference between the placebo and treatment said to be a member of their own Team “Triangle”, or of the was the absence vs. presence of the active neuropeptide. rivaling other Team “Circle” (but different from the antago- nist in the between-group competition games; we counter- 4.3. Experimental procedures and materials balanced labels but this had no effects whatsoever). There was no feedback in between rounds–participants decided to Fig. 1 provides the time-line of the experimental procedures fold, call or raise, and then immediately moved to the next and tasks. Participants came in groups of six individuals to round of the game. the laboratory, where they were seated in individual cubicles After a practice trial, participants played four rounds of in front of a computer displaying all instructions and recording poker in which we varied participants' competitive strength. responses. Participants could neither see nor communicate Participants received two low value cards (3 and 5), repre- with other participants, and worked independently and at their senting low competitive strength (the probability of winning, own pace. a draw, and losing, respectively, is 7.84%, 5.88%, and 86.27% Participants self-administered oxytocin (or placebo) under for card value 3, and 23.53%, 5.88%, and 70.59% for card value 5). experimenter supervision. Thereafter, the experimenter Participants received two high value cards (J and K), repre- unlocked the participant's computer for them to work on a senting high competitive strength (the probability of winning, series of unrelated tests, for a total duration of 30 min (this was a draw, and losing, respectively, is 70.59%, 5.88%, and 23.53% done because effects of oxytocin emerge approximately 30– for card value J, and 86.27%, 5.88%, and 7.84% for card value 35 min past administration; Baumgartner et al., 2008). Embedded K). The color of the antagonist's chips was either red or white in these unrelated tests was a measure of social value orienta- but because chip color did not interact with any of the tion, which asks participants in nine decomposed games to variables, this variable is further ignored. choose between a cooperative and a non-cooperative distribu- After the opponent placed a first bet of 10 chips, the tion of outcomes between themselves and an anonymous other participant had the choice between foldings (in which case (see for more detail, De Dreu et al., 2010). To verify that the remaining 20 chips would be preserved, but five chips randomization across experimental conditions was successful, would be lost), calling (in which case 15 chips could be won, we analyzed the number of cooperative choices (range 0–9) in a 2 or 15 chips could be lost), or raising with a maximum of 10 (Treatment) 2(Antagonist'sGroupMembership)ANOVA. chips in which case the opponent could either call the raise  Neither factor had significant effects, alone or in combination, (depending on the amount raised, up to 25 chips could be all F(1, 65)o2.34, all ps40.16, indicating that prior to the won or lost), or fold, in which case the participant would win (effective) manipulation of our independent variables, and the 15 chips. Note that the simulated poker game was an decision game itself, no differences in cooperative preference interpersonal (dyadic) game, with earnings from the game existed across the four conditions. going only to the individual participant and not (also) to other Thirty minutes past self-administration of oxytocin or members of his team and/or the (simulated) antagonist's placebo, the computer switched to instructions for the main team. Because no actual interaction took place, each partici- experimental tasks. Participants read that they were about to pant received a fixed number of chips after the four rounds of engage in a series of decision making tasks, some of which the game, and each thus had the same probability of winning would require two three-person groups. Participants were one of the raffle prizes. 178 brain research 1580 (2014) 172–179

in intergroup conflict to protect vulnerable in-group members. Acknowledgments and author contributions PlosOne 7, E46751. Declerck, C.H., Boone, C., Kiyonari, T., 2010. Oxytocin and Financial support was provided by the Affect Regulation cooperation under conditions of uncertainty: the modulating Grant of the University of Amsterdam awarded to CKWDD. role of incentives and social information. Horm. Behav. 57, 368–374. The authors declare no conflict of interest, and acknowledge Deutsch, M., 1973. The resolution of conflict: Constructive and that the funding source was not involved in the design of the destructive processes. Yale University Press, New Haven. study, the analyses, the reporting, and the decision to seek Ditzen, B., Neumann, I.D., Bodenmann, D., Von Dawans, B., publication. The authors thank Jolien van Breen and Hagar ter Turner, R.A., Ehlert, U., Heinrichs, M., 2007. Effects of different Kuilen for assistance in collecting data. FSTV, MB, SS and kinds of couple interaction on cortisol and heart rate CKWDD designed the study; FSTV and CKWDD analyzed the responses to stress in women. Psychoneuroendocrinology 32, data. FSTV, MEK, and CKWDD wrote the paper. 565–574. Domes, G., Heinrichs, M., Michel, A., Berger, C., Herpertz, S.C., 2007. Oxytocin improves “mind-reading” in humans. Biol. references Psychiatry 61, 731–733. Donaldson, Z.R., Young, L.J., 2008. Oxytocin, vasopressin, and the neurogenetics of sociality. Science 322, 900–903. Bartz, J.A., Zaki, J., Bolger, N., Hollander, E., Ludwig, N.N., Drewe, J.A., Madden, J.R., Pearce, G.P., 2009. The social network Kolevzon, A., Ochsner, K.N., 2010. Oxytocin selectively structure of a wild meerkat population: 1. Inter-group improves empathic accuracy. Psychol. Sci. 21, 1426–1428. interactions. Behav. Ecol. Sociobiol. 63, 1295–1306. Bartz, J.A., Zaki, J., Bolger, N., Ochsner, K.N., 2011. Social effects of Ebitz, R.B., Watson, K.K., Platt, M.L., 2013. Oxytocin blunts social oxytocin in humans: context and person matter. Trends Cogn. vigilance in the rhesus macaque. Proc. Natl. Acad. Sci. U.S.A Sci. 15, 301–309. 110, 11630–11635. Baumgartner, T., Heinrichs, M., Vonlanthen, A., Fischbacher, U., Feldman, R., Gordon, I., Schneiderman, I., Weisman, O., Fehr, E., 2008. Oxytocin shapes the neural circuitry of trust and Zagoory-Sharon, O., 2010. Natural variations in maternal trust adaptation in humans. Neuron 58, 639–650. and paternal care are associated with systematic changes Bos, P.A., Panksepp, J., Bluthe, R-M., Van Honk, J., 2012. Acute in oxytocin following parent-infant contact. effects of steroid hormones and neuropeptides on human Psychoneuroendocrinology 35, 1133–1141. social-emotional behavior: a review of single administration Ferguson, J.N., Young, L.J., Insel, T.R., 2002. The neuroendocrine studies. Front. Neuroendocrinol. 33, 17–35. basis of social recognition. Front. Neuroendocrinol. 23, Bosch, O.J., Meddle, S.L., Beiderbeck, D.I., Douglas, A.J., Neumann, I.D., 200–224. 2005. Brain oxytocin correlates with maternal aggression: link to Ferguson, J.N., Young, L.J., Hearn, E.F., Matzuk, M.M., Insel, T.R., anxiety. J. Neurosci. 25, 6807–6815. Winslow, J.T., 2000. Social amnesia in mice lacking the Carter, C.S., Grippo, A.J., Pournajafi-Nazarloo, H., Ruscio, M.G., oxytocin gene. Nat. Genet. 25, 284–288. Porges, S.W., 2008. Oxytocin, vasopressin, and sociality. Prog. Fischer-Shofty M., Levkovitz Y. and Shamay-Tsoory S.G., Brain Res. 170, 331–336. Oxytocin facilitates accurate perception of competition in Chang, S.W.C., Barter, J.W., Ebitz, R.B., Watson, K.K., Platt, M.L., men and kinship in women, Soc. Cog. Affect. Neurosci, 2012. Inhaled oxytocin amplifies both vicarious reinforcement http://dx.doi.org/10.1093/scan/nsr100, in press. and self reinforcement in rhesus macaques (Macace mulatta). Fischer-Shofty, M., Shamay-Tsoory, S.G., Harari, H., Levkovitz, Y., Proc. Natl. Acad. Sci. USA 109, 959–964. 2010. The effect of intranasal administration of oxytocin on Choi, J-K., Bowles, S., 2007. The coevolution of parochial altruism fear recognition. Neuropsychologia 48, 179–184. and war. Science 318, 636–640. Goodson, J.L., 2013. Deconstructing sociality, social evolution and fl De Dreu, C.K.W., 2010. Social Con ict: the Emergence and relevant nonapeptide functions. Psychoneuroendocrinology Consequences of Struggle and Negotiation. In: Fiske, S.T. (Ed.), 48, 465–478. Handbook of Social Psychology 2nd ed. Wiley, New York, pp. Gordon, I., Zagoory-Sharon, O., Leckman, J.F., Feldman, R., 2010. 983–1023. Prolactin, oxytocin, and the development of paternal behavior De Dreu, C.K.W., 2012. Oxytocin modulates cooperation within across the first six months of fatherhood. Horm. Behav. 58, and competition between groups: an integrative review and 513–518. research agenda. Horm. Behav. 61, 419–428. Heinrichs, M., Baumgartner, T., Kirschbaum, C., Ehlert, U., 2003. De Dreu C.K.W., Baas M., Roskes M., Sligte D.J., Epstein R.P., Chew Social support and oxytocin interact to suppress cortisol and S,H., Tong T., Jiang Y., Mayseless N. and Shamay-Tsoory S.G., subjective responses to psychological stress. Biol. Psychiatry. Oxytonergic circuitry sustains and enables creative cognition 54, 1389–1398. in humans, Soc. Cogn. Affect. Neurosci., http://dx.doi.org/10. Heinrichs, M., von Dawans, B., Domes, G., 2009. Oxytocin, 1093/scan/nst094, in press. vasopressin, and human social behavior. Front. De Dreu, C.K.W., Greer, L.L., Handgraaf., M.J.J., Shalvi, S., Neuroendocrinol. 30, 548–557. Van Kleef, G.A., Baas, M., Ten Velden, F.S., Van Dijk, E., Holt-Lunstad, J., Birmingham, W.A., Light, K.C., 2008. Influence of Feith, S.W.W., 2010. The neuropeptide oxytocin regulates a “warm touch” support enhancement intervention among parochial altruism in intergroup conflict among humans. married couples on ambulatory blood pressure, oxytocin, Science 328, 1408–1411. alpha amalyse, and cortisol. Psychosom. Med. 70, 976–985. De Dreu, C.K.W., Greer, L.L., Van Kleef, G.A., Shalvi, S., Hurlemann, R., Patin, A., Onur, O., Cohen, M.X., Baumgartner, T., Handgraaf, M.J.J., 2011. Oxytocin promotes human Metzler, S, Dziobek, I., Gallinat, J., Wagner, M., Kendrick, K.M., ethnocentrism. Proc. Natl. Acad. Sci. USA 108, 1262–1266. 2010. Oxytocin enhances amygdala-dependent, socially De Dreu, C.K.W., Greer, L.L., Shalvi, S., Handgraaf, M.J.J., reinforced learning and emotional empathy in humans. Van Kleef, G.A., 2012a. Oxytocin modulates the selection of J. Neurosci. 30, 4999–5007. allies in intergroup conflict. Proc. Biol. Sci 279, 1150–1154. Israel, S., Weisman, O., Ebstein, R.P., Bornstein, G., 2012. Oxytocin, De Dreu, C.K.W., Shalvi, S., Greer, L.L., Van Kleef, G.A., but not vasopressin, increases both parochial and universal Handgraaf, M.J.J., 2012b. Oxytocin motivates non-cooperation altruism. Psychoneuroendocrinology 37, 1341–1344. brain research 1580 (2014) 172–179 179

Kemp, A.H., Guastella, A.J., 2011. The role of oxytocin in human insula, and inferior frontal gyrus responses to infant crying: a affect: a novel hypothesis. Curr. Dir. Psychological Sci 20, 222–231. randomized Controlled Trial. Biol. Psychiatry 70, 291–297. Keri, S., Benedek, G., 2009. Oxytocin enhances the perception of Rimmele, U., Hediger, K., Heinrichs, M., Klaver, P., 2009. Oxytocin biological motion in humans. Cogn. Affect. Behav. Neurosci. 9, makes a face in memory more familiar. J. Neurosci. 29, 38–42. 237–241. Ross, H.E., Young, L.J., 2009. Oxytocin and the neural mechanisms Kirsch, P., Esslinger, C., Chen, Q., Mier, D., Lis, S., Siddhanti, S., regulating social cognition and affiliative behavior. Front. Gruppe, H., Mattay, V.S., Gallhofer, B., Meyer-Lindenberg, A., Neuroendocrinol. 30, 534–547. 2005. Oxytocin modulates neural circuitry for social cognition Scheele, D., Striepens, N., Gunturkun, O., Deutschlander, S., and fear in humans. J. Neurosci. 25, 11489–11493. Maier, W., Kendrick, K.M., Hurlemann, R., 2012. Oxytocin Kosfeld, M., Heinrichs., M., Zak, P.J., Fischbacher, U., Fehr, E., 2005. modulates social distance between males and females. Oxytocin increases trust in humans. Nature 435, 673–676. J. Neurosci. 32, 16074–16079. Kret, M.E., De Dreu, C.K.W., 2013. Oxytocin-motivated ally Shamay‐Tsoory, S.G., Fischer, M., Dvash, J., Harari, H., selection is moderated by fetal testosteron exposure and Perach‐Bloom, N., Levkovitz, Y., 2009. Intranasal empathic concern. Front. Neurosci 7, 1–9. administration of oxytocin increases envy and Lim, M.M., Young, L.J., 2006. Neuropeptic regulation of affiliative schadenfreude (gloating). Biol. Psychiatry. 66, 864–870. behavior and social bonding in animals. Horm. Behav. 50, Sheng, F., Liu, Y., Zhou, B., Zhou, W., Han, S., 2013. Oxytocin 506–517. modulates the racial bias in neural responses to Ludwig, M., Leng, G., 2006. Dendritic peptide release and peptide- others'suffering. Biol. Psychol. 92, 380–386. dependent behaviors. Nat. Rev. Neurosci. 7, 126–136. Singer, T., Snozzi, R., Bird, G., Petrovic, P., Silani, G., Heinrichs, M., Macbeth, A.H., Lee, H.J., Edds, J., Young III, W.S., 2009. Oxytocin Dolan, R.J., 2008. Effects of oxytocin and prosocial behavior on and the ocytocin receptor underlie intrastrain, but not brain responses to direct and vicariously experienced pain. – interstrain, social cognition. Genes Brain Behav. 8, 558–567. Emotion 8, 781 791. Skuse, D.H., Gallagher, L., 2008. Dopaminergic-neuropeptide Madden, J.R., Clutton-Brock, T.H., 2011. Experimental peripheral interactions in the social brain. Trends Cogn. Sci. 13, 27–35. administration of oxytocin elevates a suite of cooperative Stallen, M., De Dreu, C.K.W., Shalvi, S., Schmidts, A., Sanfey, A., behaviours in a wild social animal. Proc. Biol. Sci. 278, 2012. The herding hormone: oxytocin stimulates in-group 1189–1194. conformity. Psychol. Sci. 23, 1288–1292. Mikolajczak, M., Gross, J.J., Lane, A., Corneille, O., de Timary, Ph., Striepens, N., Scheele, D., Kenrick, K.M., Becker, B., Schaefer, L., Luminet, O., 2010. Oxytocin makes people trusting, not Schwalba, K., Reul, J., Maier, W., Hurlemann, R., 2012. Oxytocin gullible. Psychol. Sci. 21, 1072–1075. facilitates protective responses to aversive social stimuli in Morhenn, V.B., Park, J.W., Piper, E., Zak, P.J., 2008. Monetary sacrifice males. Proc. Natl. Acad. Sci. USA 109, 18144–18149. among strangers is mediated by endogenous oxytocin release Tatsuoka, M.M., 1988. Multivariate analysis. MacMillan, New York. after physical contact.Evol.Hum.Behav.29,375–383. Taylor, S.E., Klein, L.C., Lewis, B.P., Gruenewald, T.L., Gurung, R.A.R., Naber, F., Van IJzendoorn, M.H., Deschamps, P., Van Engeland, H., Updegraff, J.A., 2000. Biobehavioral responses to stress in females: Bakermans-Kranenburg, M.J., 2010. Intranasal oxytocin tend-and-befriend, not fight-or flight. Psych. Rev 107, 411–429. increases fathers' observed responsiveness during play with Ten Velden, F.S., Baas, M., Shalvi, S., Preenen, P.Y, De Dreu, C.K.W., their children: a double-blind within-subject experiment. 2012. In competitive interaction displays of red increase actors' – Psychoneuroendocrinology 35, 1583 1586. competitive approach and perceivers' withdrawal. Neumann, I.D., 2008. Brain oxytocin: a key regulator of emotional J. Exp. Soc. Psychol. 48, 1205–1208. and social behaviours in both females and males. Uvnas-Moberg, K., 1998. Oxytocin may mediate the benefits of J. Neuroendocrinol. 20, 858–865. positive social interaction and emotions. Parr L.A., Modi M, Siebert E. and Young L.J., Intranasal oxytocin Psychoneuroendocrinology 23, 819–835. selectively attenuates rhesus monkeys' attention to negative van Ijzendoorn, M.H., Bakermans-Kranenburg, M.J., 2012. A sniff facial expressions, Psychoneuroendocrinology, http://dx.doi. of trust: meta-analysis of the effects of intranasal oxytocin org/j.psyneuen.2013.02.011, in press. administration on face recognition, trust to in-group, and Pedersen, C.A., Ascher, J.A., Monroe, Y.L., Prange Jr., A.J., 1982. trust to outgroup. Psychoneuroendocrinology 37, 438–443. Oxytocin induces maternal behavior in virgin female rats. Veenema, A.H., Blume, A., Niederle, D., Buwalda, B., Neumann, I.D., Science 216, 648–650. 2010. Effects of early life stress on adult male aggression and Perry, A., Bentin, S., Shalev, I., Israel, S., Uzefovsky, F., hypothalamic vasopressin and serotonin. Eur. J. Neurosci. 24, Bar-On, Ebstein, R.P., 2010. Intranasal oxytocin modulates EEG 1711–1720. mu/alpha and beta rhythms during perception of biological Yoshida, M., Takaynagi, Y., Inoue, K., Kimura, T., Young, L.J., motion. Psychoneuroendocrinology 35, 1446–1453. Onaka, T., Nishimori, K., 2009. Evidence that oxytocin exerts Petrovic, P., Kalisch, R., Singer, T., Dolan, R.J., 2008. Oxytocin anxiolytic effects via oxytocin receptor expressed in attenuates affective evaluations of conditioned faces and serotonergic neurons in mice. J. Neurosci. 29, 2259–2271. amygdala activity. J. Neurosci. 28, 6607–6615. Zak, P.J., Kurzban, R., Matzner, W.T., 2005. Oxytocin is associated Riem, M.M.E., Bakermans-Kranenburg, M.J., Pieper, S., Tops, M., with human trustworthiness. Horm. Behav. 48, 522–527. Boksem, M.A.S., Vermeiren, R.R.J.M., Van IJzendoorn, M.H., Zak, P.J., Stanton, A.A., Ahmadi, S., 2007. Oxytocin increases Rombouts, S.A.R.B., 2011. Oxytocin modulates amygdala, generosity in humans. PlosOne 11, e1128. brain research 1580 (2014) 180–187

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Oxytocin eliminates the own-race bias in face recognition memory$

Iris Blando´n-Gitlina,n, Kathy Pezdekb, Sesar Saldivara, Erin Steelmana aCalifornia State University, Fullerton, USA bClaremont Graduate University, USA article info abstract

Article history: The neuropeptide Oxytocin influences a number of social behaviors, including processing Accepted 8 July 2013 of faces. We examined whether Oxytocin facilitates the processing of out-group faces and Available online 18 July 2013 reduce the own-race bias (ORB). The ORB is a robust phenomenon characterized by poor Keywords: recognition memory of other-race faces compared to the same-race faces. In Experiment 1, Oxytocin participants received intranasal solutions of Oxytocin or placebo prior to viewing White Other-race bias and Black faces. On a subsequent recognition test, whereas in the placebo condition the Face recognition same-race faces were better recognized than other-race faces, in the Oxytocin condition Encoding Black and White faces were equally well recognized, effectively eliminating the ORB. In Experiment 2, Oxytocin was administered after the study phase. The ORB resulted, but Oxytocin did not significantly reduce the effect. This study is the first to show that Oxytocin can enhance face memory of out-group members and underscore the importance of social encoding mechanisms underlying the own-race bias. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2013 Elsevier B.V. All rights reserved.

1. Introduction make faces more familiar during subsequent recognition tests, and the effect was unique to this social stimulus set; The neuropeptide Oxytocin (OT) has been implicated in recognition of non-social stimuli (e.g., houses, sculptures, and complex emotional and social behaviors. Human and animal landscapes) did not show a beneficial effect of OT. Thus, face studies show that OT enhances a number of social interac- identification may be facilitated by pre-encoding exposure to tions, reduces anxiety, increases trust, and in some species OT. In this study we examined the potential effects of this decreases social avoidance and aggression (Lee et al., 2009; peptide on the own-race bias (ORB). The ORB is a phenom- Zak, 2012). More recently the effects of OT on human social enon characterized by less accurate recognition of other-race cognition have been demonstrated as well (for reviews see than the same-race faces due to encoding failures. We predict Bartz et al., 2011; MacDonald and MacDonald, 2010). Rimmele that pre-encoding OT exposure will lead to improvements in et al. (2009) reported that pre-encoding intranasal doses of OT recognition of other-race faces.

☆This research was supported by seed Grants to the first author from California State University, Fullerton. Part of this research served as partial fulfillment of Erin Steelman′s honors thesis supported by the California State Fullerton MARC UnSTAR Program Grant 2T34GM008612-15 and -16 from the National Institutes of Health. nCorrespondence to: CSUF, Department of Psychology, P.O. Box 6846, Fullerton, CA 92834-6846, USA. Fax: +1 657 278 7134. E-mail address: [email protected] (I. Blandón-Gitlin).

0006-8993/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.brainres.2013.07.015 brain research 1580 (2014) 180–187 181

1.1. Social-cognitive factors underlying the own-race bias Although the mechanisms by which OT facilitates face recognition are unknown, it has been suggested that OT may The ORB is a robust phenomenon demonstrated in numerous work by reducing social anxiety and stress via attenuation of studies in the lab and real-world settings (Brigham et al., amygdala activity, which induces autonomic responses to 2007). In a typical ORB study, participants view the same- and emotional experience (Kirsch et al., 2005). Using functional other-race faces. On subsequent recognition memory tests, MRI, Domes et al. (2007a) showed that compared to a placebo the same-race faces are more accurately recognized than group, OT-treated males had less right-sided amygdala activ- other-race faces by a number of groups (see Brigham et al., ity in response to emotional face stimuli (showing angry, 2007; Pezdek et al., 2003). A principal explanation of the ORB fearful or happy expressions) than neutral stimuli. posits that the effect results from social-cognitive factors that Another mechanism by which OT may enhance face lead to differential processing of the same- and other-race memory is by facilitating eye-gaze to face regions conveying faces (see Levin, 2000; Meissner et al., 2005; Sporer, 2001). important social information. Guastella et al. (2008a) had When a face is perceived, an automatic social categorization participants in OT and placebo conditions view neutral faces. process occurs such that the face is placed into an in-group or They recorded eye-gaze duration and fixation toward three an out-group category. Race is one of a number of factors that face regions: eye, nose/mouth and forehead/cheek. Based on can cue out-group categorization processes (Levin, 2000; Susa previous research indicating that the eye region is primary in et al., 2010). As demonstrated by studies with ambiguous race social face perception, the authors hypothesized this region faces, changing just one facial feature that stereotypically would be important in eye-gaze patterns. Compared to the marks a race (e.g. hair style) can trigger face categorization placebo group, OT-treated participants gazed significantly processes (Maclin and Malpass, 2001). This categorization longer and more frequently at the eye region of faces occurs with other in-group/out-group social categories as compared to other regions. These eye-gaze data, together well. Shriver et al. (2008) showed that White participants with the face recognition studies, suggest that OT may remembered faces of White individuals more accurately if facilitate encoding by inducing a face individuation orienting they believed them to be affiliated with their own rather than response. another university or social economic class. Cognitively, the categorization of faces can be character- 1.3. Influence of Oxytocin on social categorization ized as an economy of attentional resources. Whereas out- group faces are “cognitively disregarded” (Rodin, 1987) lead- Research investigating the influence of OT on social categor- ing to decreased attention and shallow processing, in-group ization in humans is sparse (de Dreu, 2012). However, recent faces receive more attentional resources to individuating research shows that OT may moderate socially relevant features diagnostic of recognition (Meissner et al., 2005). behaviors toward in-group and to some extent out-group Indeed instructing participants to pay close attention to members. Using measures of event-related brain potentials, individual aspects of other-race faces substantially reduces Sheng et al. (2013) found greater neural response to in-group the ORB (Hugenberg et al., 2007). Lebrecht et al. (2009) found faces displaying pain expressions than neutral expressions. that training involving an individuation task (attending to This empathy bias was enhanced by OT-treatment signifi- face differentiating cues) but not category tasks (deciding on cantly more when viewing in-group faces than out-group race of face) reduced both the ORB and implicit racial bias. faces. Similarly, de Dreu et al. (2011) reported that OT-treated Thus, it appears that early processing dimensions of social groups made stronger positive implicit associations (e.g., categorization affect attentional allocation and to some ascribing words to targets) and benevolent decisions (e.g., extent attitudes toward out-group members. These mechan- reduced readiness to sacrifice targets) about the in-group isms appear to underlie the bias in other-race face recogni- compared to control conditions. Only in some conditions did tion memory. OT-treated participants make negative implicit associations toward the out-group, a result that has not been specifically clarified and may be explained by out-group derogation (de 1.2. Influence of Oxytocin on face processing Dreu et al. 2011) or a general OT-induced sensitivity to socially important information (Averbeck, 2010; Chen et al., In humans, the impact of OT on social recognition and 2011). In these social categorization studies OT had a greater behavior appears to be moderated by contextual and indivi- effect on attitudes and behaviors toward in-group than out- dual difference factors (Bartz et al., 2011). Despite this, a group members. The potential effect of OT on face encoding recent meta-analysis of the effect of OT on facial recognition of and recognition memory for in-group vs. out-group members emotions showed a modest significant effect size in the remains unexplored. Examining this relationship is a princi- direction of OT-treated groups outperforming the placebo pal goal of the current study. group (overall d .21, po0.01; Van IJzendoorn and ¼ Bakermans-Kranenburg, 2012). Similarly, four of the five 1.4. Current study published studies on face recognition memory show positive effects in the predicted direction (Di Simplicio et al., 2009; Given that OT can facilitate encoding of faces via a combina- Guastella et al., 2008b; Rimmele et al., 2009; Savaskan et al., tion of social, neurological and cognitive processes, and the 2008). A reverse effect was reported by Herzmann et al. (2012). finding that the ORB results from encoding deficits for other- Thus, the overall pattern of result shows that OT enhances race faces in some of these same processes, the principal face processing and recognition. prediction in this study is that encoding of other-race faces 182 brain research 1580 (2014) 180–187

will be enhanced by the application of OT compared to a Hit Rate False Alarm Rate d' Rate C Rate placebo control condition. In two double-blind experiments, 2 1.75 participants were randomly assigned to an OT or placebo 1.5 condition. Using a social orienting task they encoded own- 1.25 1 and other-races faces followed by a recognition memory test. 0.75 In Experiment 1, OT was administered prior to studying the 0.5 0.25 faces to assess the effects of OT at encoding. We predict that 0 -0.25 OT will enhance face recognition overall. More important, we Placebo OT Placebo OT Placebo OT Placebo OT predict an interaction with greater improvements in recogni- Condition tion accuracy in the OT than the placebo condition for other- White Faces race than the same-race faces, effectively reducing the ORB. Black Faces The application of OT is predicted to reduce the face categor- Fig. 1 – Experiment 1 (pre-encoding substance intake). Four ization bias thereby resulting in more similar allocation of dependent measures presented as a function of race of face attentional resources to the same- and other-race faces. In and condition. Errors bars represent standard errors of the Experiment 2, OT was administered after encoding faces to means. The principal finding on the d′ rate shows the determine if OT affects the consolidation of the information predicted interaction; whereas in the placebo condition the processing stage. same-race faces were better recognized than other-race faces, in the Oxytocin condition the same- and other-race faces were equally well recognized, effectively eliminating 2. Experiment 1 results and discussion the own-race bias.

This study involves a 2 (Race of face: White or Black) by 2 (Condition: Placebo or Oxytocin), mixed design, with condi- Overall White faces (M 1.89, SE 0.09) were better recognized ¼ ¼ tion as the between-subjects factor. Measures from Signal than Black faces (M 1.56, SE 0.08), F (1, 41) 14.22, p 0.001, ¼ ¼ ¼ ¼ Detection Theory (SDT; see Banks, 1970; Stanislaw and η2 0.26. The significant interaction, F (1, 41) 4.34, p 0.044, p ¼ ¼ ¼ Todorov, 1999) were used to assess recognition accuracy for η2 0.10, showed that whereas in the placebo condition, as p ¼ the same- and other-race faces (see Section 6 for explanation predicted, White faces (M 1.90, SE 0.13) were better recog- ¼ ¼ of stimuli and participants′ tasks). The SDT measures were nized than Black faces (M 1.39, SE 0.12); in the OT condition ¼ ¼ calculated by comparing test responses to previously seen recognition accuracy for White faces (M 1.89, SE 0.12) and ¼ ¼ faces (“old”) and distractor faces presented only at test Black faces (M 1.74, SE 0.11) did not differ. It was specifi- ¼ ¼ (“new”). A hit rate indicates the proportion of correct cally recognition memory for Black faces that was improved responses to previously seen faces; a false alarm rate indi- in the OT compared to the control condition, t (41) 2.19, ¼À cates the proportion of incorrect responses to new faces. The p 0.034, d 0.68. No other effect was statistically significant ¼ ¼À d' rate is the standardized difference in accuracy between on d′ (highest F 1.30) ¼ responses to old and new faces. The C rate, also known as The 2 2 mixed ANOVAs on hits and false alarm data  response bias, is a measure of the tendency to respond showed statistically significant effects of race of face only. Hit “old” or “new” independent of accuracy. A C rate of zero rates were significantly lower for White (M 0.76, SE 0.02) ¼ ¼ indicates relatively equal distributions of “old” and “new” than Black faces (M 0.81, SE 0.02), F (1, 41) 6.81, p 0.013, ¼ ¼ ¼ ¼ responses and therefore no bias in responding. A C rate η2 0.14. However, false alarms were significantly higher with p ¼ smaller than zero indicates more “old” responses, and a C Black (M 0.28, SE 0.02) than White faces (M 0.17, SE 0.02), ¼ ¼ ¼ ¼ “ ” 2 fi rate larger than zero indicates more new responses. A lower F (1, 41) 37.12, po0.001, ηp 0.48. The nding that the ORB ¼ ¼ C rate is interpreted as a liberal response bias (more likely to was more evident for false alarm than hit rate data is respond “old” to both old and new faces), and a higher C rate consistent with findings reported elsewhere (e.g., Meissner is interpreted as a conservative bias (less likely to respond and Brigham, 2001). No other effect was statistically signifi- “old”). cant (highest F 2.44). ¼ All statistical tests in this study were conservatively A2 2 mixed ANOVA on C rates revealed a significant  conducted as two-tailed. Degrees of freedom for tests vary effect of race of face only. Participants were more liberal because some participants did not provide responses in one (more likely to respond “old” to both old and new faces) in or more conditions. Analyses of face recognition measures their judgment with Black (M 0.05, SE 0.07) than White ¼À ¼ are reported first, followed by analyses of the social encoding faces (M 0.16, SE 0.06), F (1, 41) 12.48, p 0.001, η2 0.23. No ¼ ¼ ¼ ¼ p ¼ task and self-reported perception of anxiety, mood and other effect was significant (Fo1). attention. A multivariate analysis of confidence ratings in response to the four types of stimuli at test (old-White, old-Black, new- 2.1. Face recognition White, new-Black) as a function of condition revealed no significant effects, Fo1. Fig. 1 presents the mean hit rate, false alarm rate, d′ rate and C rate as a function of race of target and condition. On the d′ 2.2. Social engagement during encoding data a 2 (Face: White vs. Black) 2 (Condition: Control vs. OT)  mixed ANOVA revealed a significant main effect of race of A2 2mixedANOVAonstimulusapproachabilityratings  face and more important, an interaction with condition. during encoding showed no statistically significant effects, Fo1. brain research 1580 (2014) 180–187 183

The overall mean approachability rating was relatively low was similar to that used in Experiment 1 except that the OT (scale 1–10, M 4.29, SE 0.22), and in all conditions none of and placebo were administered after presentation of the ¼ ¼ the correlations between face approachability rating and d′ target faces. were significant (p40.05). It is not clear why these results were not significant. Perhaps the task was only effective orienting participants to the faces but not sensitive enough 4. Experiment 2 results and discussion to determine other relationships. Experiment 2 applies the same analytical procedures from 2.3. Wakefulness, calmness, mood, and attention Experiment 1.

Two 2 (Condition) 2 (Time: pre- vs. post-encoding) mixed  4.1. Face recognition ANOVAs on wakefulness and calmness ratings were con- ducted. A significant interaction on calmness ratings Fig. 2 presents mean hit rate, false alarm rate, d′ rate, and C rate resulted, F (1,35) 8.37, p 0.007, η2 0.19; whereas in the ¼ ¼ p ¼ as a function of the race of target and condition. Various 2 control condition calmness ratings were lower pre-test (Face: White vs. Black) 2 (Condition: Control vs. OT) mixed (M 41.44, SE 1.58) than post test (M 44.94, SE 1.83), (t  ¼ ¼ ¼ ¼ ANOVAs on each measure showed statistically significant (17) 2.81, p 0.012 d 0.482), in the OT condition pre- ¼À ¼ ¼À effects of race of face only; The d′ analysis showed that White (M 44.63, SE 1.19) and post (M 43.95, SE 1.17) calmness ¼ ¼ ¼ ¼ faces (M 1.68, SE 0.11) were better recognized than Black ratings did not significantly differ, (to1). No other effect was ¼ ¼ faces (M 1.24, SE 0.09), F (1, 42) 10.58, p 0.002, η2 0.20. statistically significant (highest F 3.79). ¼ ¼ ¼ ¼ p¼ ¼ Consistent with Experiment 1, whereas hit rates were lower for The 2 2 ANOVAs on both affect ratings revealed a main  White (M .71, SE 0.02) than Black faces (M 0.76, SE 0.02), F effect of time on negative affect only. Negative affect ratings ¼ ¼ ¼ ¼ (1, 42) 6.99, p 0.011, η2 0.14, false alarms were significantly given prior to substance intake (M 15.58, SE 0.78) decreased ¼ ¼ p¼ ¼ ¼ higher with Black (M 0.34, SE 0.03) than White faces after substance intake (M 14.10, SE 0.78), F (1,41) 7.62, ¼ ¼ 2 ¼ ¼ ¼ (M 0.18, SE 0.02), F (1, 42) 51.17, po0.001, ηp 0.55. No other p 0.009, η2 0.16. No other effect was significant (highest ¼ ¼ ¼ ¼ ¼ p ¼ effect on these three measures was statistically significant F 1.74). Additionally, no significant effects were found for ¼ (highest F 2.57). Most critically, the interaction of condition the attention measure, Fo1. Together these results suggest ¼ by race of face was not significant on any of the recognition that for the most part OT did not negatively affect partici- accuracy measures including d′ rate (F 1.50, p 0.227). pants' mood, anxiety or attention. ¼ ¼ A2 2 mixed ANOVA on C rates revealed significant main In Experiment 1, the principal analyses involving discri-  effects of face and condition. Participants were more liberal in mination accuracy (d′) revealed three important findings. their judgment of Black (M 0.13, SE 0.06) than White faces First, the predicted interaction was confirmed; whereas in ¼À ¼ (M 0.28, SE 0.06), F (1, 42) 39.26, po0.001, η2 0.48, and the placebo condition same-race faces were better recognized ¼ ¼ ¼ p ¼ participants were more conservative (less likely to choose than other-race faces, in the Oxytocin condition Black and “old”) in the OT (M 0.19, SE 0.07) than placebo condition White faces were equally well recognized, effectively elim- ¼ ¼ (M 0.03, SE 0.07), F (1, 42) 4.81, p 0.034, η2 0.10. inating the ORB. Second, the main effect of OT was not ¼À ¼ ¼ ¼ p ¼ As in Experiment 1, a multivariate analysis of confidence significant. Overall better face recognition with OT was ratings in response to the four types of stimuli at test, as a predicted from previous findings but did not result. Third, function of condition was not significant, F 1.37. there was an overall ORB effect; the same-race faces were ¼ more accurately recognized than other-race faces. Finally, OT did not adversely affect participants′ anxiety, mood or Hit Rate False Alarm Rate d' Rate C Rate attention. 1.9 1.65 1.4 1.15 3. Experiment 2 0.9 0.65 0.4 In Experiment 1, pre-encoding of OT significantly improved 0.15 memory for other-race faces, thus confirming that OT -0.1 -0.35 improves face recognition memory at the encoding stage. Placebo OT Placebo OT Placebo OT Placebo OT However, Savaskan et al. (2008) reported that post-encoding Condition of OT also improved face recognition memory, thus suggest- White Faces Black Faces ing that OT may affect memory consolidation as well as encoding. The results of Experiment 1 and those of Savaskan Fig. 2 – Experiment 2 (post-encoding substance intake). Four et al. suggest that pre- and post-encoding OT administration dependent measures presented as a function race of face may differentially affect various aspects of face recognition and condition. Errors bars represent standard errors of the memory. Experiment 2 was conducted to assess if OT affected means. The principal finding on the d′ rate shows a main the own race bias at the consolidation stage in addition to the effect of face but not a significant interaction. This indicates encoding stage. If so, the significant interaction of race by that the typical own-race bias occurred in both treatment condition on d′ data reported in Experiment 1 would be conditions and post-encoding OT did not have a significant replicated in Experiment 2. The procedure of Experiment 2 effect on recognition memory. 184 brain research 1580 (2014) 180–187

4.2. Social engagement during encoding involved the significant interaction of condition by race of face in Experiment 1; compared to the placebo condition, pre- A2 2 mixed ANOVA on stimulus approachability ratings encoding intake of OT led to a significant increase in accuracy  during encoding showed a significant main effect of race of recognizing other-race faces. This interaction was not sig- face despite an overall low mean rating (scale 1–10, M 4.55, nificant in Experiment 2 when OT was administered post- ¼ SE 0.21). White faces received higher approachability ratings encoding of faces. These results suggest that the effect of OT ¼ (M 4.75, SE 0.23) than Black faces (M 4.36, SE 0.23), F (1, on other-race face processing occurs primarily at the encod- ¼ ¼ ¼ ¼ 42) 5.75, p 0.011, η2 0.12. The correlations between ing stage and does not affect memory consolidation. These ¼ ¼ p ¼ approachability ratings and d′ measures were small; none results are consistent with social-cognitive models of the ORB were significant, p40.05. and in line with many behavioral studies demonstrating that the memory impairment of other-race faces occurs because 4.3. Wakefulness, calmness, mood, and attention of encoding failures. Contrary to predictions from the OT face recognition research, there was no significant effect of OT on Two 2 (Condition) 2 (Time: pre- vs. post-encoding) mixed recognition memory for the same-race faces.  ANOVAs on wakefulness and calmness ratings were con- In modeling the basic mechanisms underlying the ORB, ducted. Only a significant main effect of wakefulness Meissner et al. (2005) and more recently Susa et al. (2010) resulted; pre-encoding ratings were higher (M 43.31, concluded that the cognitive processes involved in encoding ¼ SE 1.15) than post-encoding ratings (M 41.28, SE 1.32), F the same-race faces are qualitatively different from that of ¼ ¼ ¼ (1,39) 5.42, p 0.025, η2 0.12. No other effect was significant other-race faces. Racial categorization plays a significant role ¼ ¼ p ¼ (highest F 1.35). in early stages of other-race face processing whereby an out- ¼ The 2 2 ANOVAs on both affect ratings revealed a main group label may “disrupt the successful encoding of individ-  effect of Time on negative affect only. Negative affect ratings uating facial information” (p. 535). With the same-race faces a declined from pre-encoding (M 15.42, SE 0.94) to post- deeper processing involving the encoding of individuating ¼ ¼ encoding (M 13.78, SE 0.70), F (1,41) 6.13, p 0.018, features diagnostic of recognition occurs and can result in ¼ ¼ ¼ ¼ η2 0.13. No other effect was significant (highest F 1.27) better memory. Our results are consistent with a differential p ¼ ¼ including results on the attention measure, Fo1. processing of faces framework of the ORB. Given that OT In Experiment 2 the principal analyses involving discrimi- facilitates attention to more socially significant face regions nation accuracy (d′) revealed a significant ORB effect only. such as the eyes, it is possible that OT induced an orienting Unlike Experiment 1, in Experiment 2, when OT was admi- response to individuating encoding of other-race faces in our nistered post-encoding, the interaction of condition by race of participants. Di Simplicio et al. (2009) reported that OT face was not significant. This nonsignificant interaction in induced a slower processing speed for fearful facial expres- Experiment 2 suggests that OT affects recognition memory sions compared to other basic expressions. Thus it is possible for other-race faces only when administered prior to encod- that OT works to facilitate greater attention allocation to ing. This is consistent with social-cognitive models of the particular aspects of important stimuli and thereby strength- ORB and underscores the importance of encoding mechan- ening encoding. isms involved in the ORB effect. Emotions may play an important role as well. The emo- The main difference between Experiment 1 and 2 is the tional salience of Black faces to White participants was timing of the substance administration relative to the study shown by Cunningham et al. (2004) with functional MRI. phase. This procedure separates the OT encoding effects from These researchers showed that during a fast-pace perception consolidation effects. The effect of OT on consolidation interval, amygdala activity increased more in response to processes independent of retrieval processes cannot be Black than White faces, and this difference in activation was assessed in this study. It is likely that the OT substance was stronger in participants who exhibited greater racial bias still present in the central nervous system at the time of toward out-group members. Because OT dampens amygdala the recognition memory test, as it takes a few hours to wash activity in males (Domes et al., 2007a; Kirsch et al., 2005) it is out of the body. Several studies have used longer delays possible that in our study, OT attenuated amygdala activity in (e.g., 24 h) for the recognition test (e.g., Guastella et al., 2008b; response to emotion likely triggered by Black faces. This is Rimmele et al., 2009), and Savaskan et al. (2008) used a short possible given that the social group for most of our partici- (30 min) vs. long (24 h) delay to assess potential interactive pants involved mainly Whites. Further research is needed to effects of OT with time intervals. Results from these studies directly assess these potential relationships and importantly, showed no significant interactive effects, suggesting that OT female participants should be included. A different pattern of did not directly affect retrieval processes. Despite this, it is results may emerge or elucidate on the current findings, as in possible that in the context of our study the lack of significant some cases female participants′ amygdale activation differs main effect of OT on face recognition memory may be related from that of men. For example, Domes et al. (2010) found that to the short delay interval. compared to a control group, OT-treated females showed an increase in left amygdala activation in response to faces displaying a fearful expression. This increase in amygdala 5. General discussion activation in females followed by OT treatment may reveal additional relationships between OT and the ORB. Perhaps This study is the first to show that the neuropeptide, the ORB will increase in females due to OT-induced activation Oxytocin, reduced the own-race bias. The principal finding of the amygdala. Alternatively, gender might moderate the brain research 1580 (2014) 180–187 185

relationship between OT and ORB when faces displaying Phase 1 Phase 2 Phase 3 emotions are shown, but not when neutral faces (like the 0 min. 150 min. ones in the current study) are shown. Any of these findings 45 min.

would further our understanding of OT effects on the ORB. OT/Saline uptake period Together, these explanations help clarify the OT effect on Experiment 1 memory for other-race faces, however, they do not account 1 2 3 4 5 6 7 8 for the absence of an effect of OT on the same-race faces. Several contextual and individual difference factors have Experiment 2 fi 45 min. been noted to account for some inconsistent ndings in uptake period studies of OT and social cognition (Bartz et al., 2011). One of fi those factors is task dif culty. Domes et al. (2007b) reported OT/Saline Events and Behavioral Tasks: that participants in an OT condition performed better than 1. Health questionnaire and BP/Heart Rate reading. controls on a difficult version of a mind reading task, how- 2. Psychological processes questionnaires. 3. Exp 1: OT or Saline administration. Exp 2: Study face stimuli. ever, the two groups performed equally well on the easy 4. Exp 1: Study face stimuli. Exp 2: OT or Saline administration. 5. Psychological processes questionnaires. version of this task. Similarly, in the verbal memory domain, 6. Test of face stimuli. 7. Contact and Knowledge questionnaires. Heinrichs et al. (2004) showed that pre-encoding OT signifi- 8. Debrief. cantly impaired memory when the recall task was difficult Fig. 3 – Events and behavioral tasks completed in each phase (word generation) but not when it was easy (word-stem of Experiments 1 and 2. completion). Together these findings suggest that stimuli type, task type, and task difficulty may moderate the effect of OT on the same-and cross-race faces. Specifically, it is possible that in our study, OT had a greater influence on the design is a 2 (Race of face: White or Black) by 2 (Condition: more difficult other-race face memory task but not on the Placebo or Oxytocin), mixed design, with condition as the relatively easier same-race face memory task. between-subjects factor. This study demonstrated that Oxytocin can enhance the encoding of out-group faces, a class of stimuli that typically 6.2. Materials and procedure receives less attention, is “cognitively disregarded” and pro- cessed at a shallow level. The mechanisms underlying these There were three phases, illustrated in Fig. 3.Inphase1, encoding effects need to be clarified in future studies, participants completed a questionnaire assessing their prior including the investigation of possible motivational factors days′ activities, followed by assessments of heart rate and blood driving attention allocation to out-group faces. This study pressure. Next, questionnaires assessing wakefulness, calmness also demonstrates the feasibility of a new methodology to (Steyer et al., 1997), mood (Positive and Negative Affect Scale, test models of the ORB. Imaging studies have shed light on PANAS; http://ir.uiowa.edu/psychology_pubs/11)andattention the neural structures associated with the same- vs. other- (Brickenkamp and Zillmer, 1998)werecompletedself-paced. race face processing, behavioral studies have investigated the Immediately afterward, participants self-administered intrana social-cognitive mechanisms involved in processing these sally, either 24 IU of Oxytocin (Syntocinon Spray by Novartis; two categories of faces, and the current study contributes to three puffs of 4-IU in each nostril) or the placebo (saline) solution. the understanding of the ORB from a neurochemical perspec- Using procedures similar to Rimmele et al. (2009) we tive. This furthers our understanding of the ORB and provides allowed 45 min to ensure OT central nervous system effects. a more complete picture of its underlying mechanisms. During this time participants completed distractor tasks. In phase 2, participants viewed 50 neutral-expression faces (25 Black, 25 White, in random blocks) for 2.5 s each, with instruc- 6. Experiment 1 method tions to study each face and indicate the extent to which they would like to approach the person (rating: 1 not at all; ¼ 6.1. Participants and design 10 very much). This orienting task (a) ensured that partici- ¼ pants were attending to stimuli in a social manner, and (b) Forty-three non-Hispanic Caucasian males (age M 19.86, determine if OT has a greater effect than placebo on subjective ¼ SD 1.68) meeting criteria were included. They received $20 perception of stimuli at encoding. Practice trials familiarized ¼ and course credit. Similar to other studies (e.g., Rimmele participants with the task. The face stimuli for both experi- et al., 2009; Savaskan et al., 2008) participants′ inclusion ments were selected from the database of male faces with criteria were: no physical or mental illness, not taking pre- neutral expressions used by Meissner et al. (2005) and others.1 scription medication, non-smokers, and a normal heart rate Immediately following, in phase 3, participants again and blood pressure at the time of the study. Participants completed the wakefulness, calmness, mood and attention reported normal sleep–wake cycle the day preceding the questionnaires. They were then instructed on the test proce- study, and no heavy consumption of alcohol or caffeinated dure and given practice trials. The test consisted of 50 old and drinks, or use of illegal drugs prior to testing. Three partici- 50 new face stimuli (50 Black and 50 White faces) presented pants who showed deviant scores on hits or false alarm for 2.5 s each and organized in blocks shown in random order. measures (z43.29) were not included in analyses. A double- Participants′ task was to decide if each face was old or new blind placebo-control design was used with 23 participants in the OT condition and 20 in the control condition. The main 1See: http://iilab.utep.edu/stimuli.htm. 186 brain research 1580 (2014) 180–187

and rate their confidence in that assessment (1 not at all de Dreu, C.K.W., 2012. Oxytocin modulates cooperation within ¼ confident; 10 very confident). and competition between groups: an integrative review and ¼ Participants then reported their contact with the same- research agenda. Horm. Behav. 61 (3), 419–428, http://dx.doi. org/10.1016/j.yhbeh.2011.12.009. and other- race groups; the majority (73%) reported that their de Dreu, C.K.W., Greer, L.L., Van Kleef, G.A., Shalvi, S., Handgraaf, social group included mostly Whites. Finally, they were asked M.J.J., 2011. Oxytocin promoted human ethnocentrism. Proc. to guess which condition they were assigned to. Only 17 Natl. Acad. Sci. USA 108, 1262–1266, http://dx.doi.org/10.1073/ provided a guess response. The proportion of those guessing pnas.1015316108. correctly (47%) was not significantly different from those Di Simplicio, M., Massey-Chase, R., Cowen, P., Harmer, C., 2009. guessing incorrectly (54%), Z .34, p40.05. Oxytocin enhances processing of positive vs. negative ¼ emotional information in healthy male volunteers. J. Psychopharmacol. 23, 241–248, http://dx.doi.org/10.1177/ 7. Experiment 2 method 0269881108095705. Domes, G., Heinrichs, M., Gläscher, J., Büchel, C., Braus, D.F., 7.1. Participants and design Herpertz, S., 2007a. Oxytocin attenuates amygdala response to emotional faces regardless of valence. Biol. Psychiatry 62, 1187–1190, http://dx.doi.org/10.1016/j.biopsych.2007.03.025. Forty-four non-Hispanic Caucasian males (age M 19.57, ¼ Domes, G., Heinrichs, M., Michel, A., Berger, C., Herpertz, S., SD 1.45) participated. They received $20 and course credit. “ ” ¼ 2007b. Oxytocin improves mind-reading in humans. Biol. The same design, inclusion criteria, and prescreening used in Psychiatry 61, 731–733, http://dx.doi.org/10.1016/j. Experiment 1 were used in Experiment 2. biopsych.2006.07.015. Domes, G., Lischke, A., Berger, C., Grossmann, A., Hauenstein, K, 7.2. Materials and procedure Heinrichs, M., Herpertz, S.C., 2010. Effects of intranasal oxytocin on emotional face processing in women. Psychoneuroendocrinology 35, 83–93, http://dx.doi.org/ In phase 1, all procedures followed in Experiment 1 were used 10.1016/j.psyneuen.2009.06.016. except that participants studied the face stimuli in phase 1 Guastella, A.J., Mitchell, P.B., Dadds, M.R., 2008a. Oxytocin rather than phase 2 (see Fig. 3). Similar to Savaskan et al. increases gaze to the eye-region of human faces. Biol. (2008), immediately following presentation phase, in phase 2, Psychiatry 63, 3–5, http://dx.doi.org/10.1016/j. participants self-administered the substance and waited biopsych.2007.06.026. 45 min. In phase 3, participants again completed the wake- Guastella, A.J., Mitchell, P.B., Mathews, F., 2008b. Oxytocin enhances the encoding of positive social memories in fulness, calmness, mood and attention questionnaires and humans. Biol. Psychiatry 64, 256–258, http://dx.doi.org/ were then presented the face recognition memory test. 10.1016/j.biopsych.2008.02.008. Next participants reported on their contact with the same- Heinrichs, M., Meinlschmidt, G., Wippich, W., Ehlert, U., and other-race groups; 63% reported their social group to Hellhammer, D.H., 2004. Selective amnesic effects of oxytocin include mostly Whites. Finally participants were asked to on human memory. Physiol. Behav. 83, 31–38, http://dx.doi. guess which condition they were assigned to. Only 20 org/10.1016/j.physbeh.2004.07.020. participants provided a guess response. The proportion of Herzmann, G., Young, B., Bird, C.W., Curran, T., 2012. Oxytocin those guessing correctly (35%) was not significantly different can impair memory for social and non-social visual objects: a within-subject investigation of oxytocin′s effects on human from those guessing incorrectly (65%), Z 1.90, p40.05 ¼ memory. Brain Res. 1451, 65–73, http://dx.doi.org/10.1016/j. brainres.2012.02.049. references Hugenberg, K., Miller, J., Claypool, H., 2007. Categorization and individuation in the cross- race recognition deficit: toward a solution to an insidious problem. J. Exp. Soc. Psychol. 43, Averbeck, B.B., 2010. Oxytocin and the salience of social cues. 334–340, http://dx.doi.org/10.1111/j.1751-9004.2008.00090.x. Proc. Natl. Acad. Sci. USA 107, 9033–9034, http://dx.doi.org/ Kirsch, P., Esslinger, C., Chen, Q., Mier, D., Lis, S., Siddhanti, S., et al., 10.1073/pnas.1004892107. 2005. Oxytocin modulates neural circuitry for social cognition Banks, W.P., 1970. Signal detection theory and human memory. and fear in humans. J. Neurosci. 25, 11489–11493, http://dx.doi. Psychol. Bull. 74, 81–99, http://dx.doi.org/10.1037/h0029531. org/10.1523/JNEUROSCI.3984-05.2005. Bartz, J.A., Zaki, J., Bolger, N., Ochsner, K.N., 2011. Social effects of Lebrecht, S., Pierce, L., Tarr, M.J., Tanaka, J., 2009. Perceptual oxytocin in humans: context and person matter. Trends Cogn. other-race training reduces implicit racial bias. PLoS One 4 (1), Sci. 15, 301–309, http://dx.doi.org/10.1016/j.tics.2011.05.002. e4215, http://dx.doi.org/10.1371/journal.pone.0004215. Brigham, J.C., Bennett, L.B., Meissner, C.A., Mitchell, T., 2007. The Lee, H.-J., Macbeth, A.H., Pagani, J.H., Young, W.S., 2009. Oxytocin: influence of race on eyewitness memory. In: Lindsay, R., Ross, the great facilitator of life. Prog. Neurobiol. 88, 127–151, http: D., Read, J., Toglia, M. (Eds.), The Handbook of Eyewitness //dx.doi.org/10.1016/j.pneurobio.2009.04.001. Psychology: Memory for People. Lawrence Erlbaum Associates Levin, D.T., 2000. Race as a visual feature: using visual search and Publishers, Mahwah, NJ, US, pp. 257–281. perceptual discrimination tasks to understand face categories Brickenkamp, R., Zillmer, E.A., 1998. d2 Test of Attention. Hogrefe and the cross-race recognition deficit. J. Exp. Psychol. 129, & Huber, Göttingen, Germany. 559–574, 10.1037S0096-3445.129.4.559. Chen, F.S., Kumsta, R., Heinrichs, M., 2011. Oxytocin and Meissner, C.A., Brigham, J.C., Butz, D.A., 2005. Memory for own- intergroup relations: goodwill is not a fixed pie. Proc Natl. and other-race faces: a dual-process approach. Appl. Cogn. Acad. Sci 108, E45, http://dx.doi.org/10.1073/pnas.1101633108. Psychol. 19, 545–567, http://dx.doi.org/10.1002/acp.1097. Cunningham, W.A., Johnson, M.K., Raye, C.L., Gatenby, J.C., Gore, Meissner, C.A., Brigham, J.C., 2001. Thirty years of investigating J.C., Banaji, M.R., 2004. Separable neural components in the own-race bias in memory for faces: a meta-analytic the processing of Black and White faces. Psychol. Sci. 15, review. Psychol. Public Policy Law 7, 3–35, http://dx.doi.org/ 806–813. 10.1037//1076-8971.7.1.3. brain research 1580 (2014) 180–187 187

MacDonald, K., MacDonald, T.M., 2010. The peptide that binds: a Shriver, E.R., Young, S.G., Hugenberg, K., Bernstein, M.J., Lanter, J. systematic review of oxytocin and its prosocial effects in R., 2008. Class, race, and the face: Social context modulates humans. Harv. Rev. Psychiatry 18, 1–21, http://dx.doi.org/ the Cross-Race Effect in face recognition. Pers. Soc. Psychol. 10.3109/10673220903523615. Bull. 34, 260–278, http://dx.doi.org/10.1177/0146167207310455. Maclin, O.H., Malpass, R.S., 2001. Racial categorization of faces: Sporer, S.L., 2001. Recognizing faces of other ethnic groups: an the ambiguous race face effect. Psychol. Public Policy Law 98, integration of theories. Psychol. Public Policy Law 7, 36–97, 112–113, http://dx.doi.org/10.1037//1076-8971.7.1.98. http://dx.doi.org/10.1037//1076-8971.7.136. Pezdek, K., Blandón-Gitlin, I., Moore, C., 2003. Children′s face Stanislaw, H., Todorov, N., 1999. Calculation of signal detection recognition memory: more evidence for the Cross-Race Effect. theory measures. Behav. Res. Methods Instrum. Comput. 31, 137–149, http://dx.doi.org/10.3758/BF03207704. J. Appl. Psychol. 88, 760–763, http://dx.doi.org/10.1037/0021- Steyer, R., Schwenkmezger, P., Notz, P., Eid, M., 1997. Der 9010.88.4.760. Mehrdimensionale Befindlichkeitsbogen MDBF. Hogrefe, Rimmele, U., Hediger, K, Heinrichs, M., Klaver, P., 2009. Oxytocin Göttingen. Retrieved from University of Jena, Professor Steyer′s makes a face in memory familiar. J. Neurosci. 29, 38–42, http: Web site: 〈http:/www.metheval.uni-jena.de/mdbf.php〉. //dx.doi.org/10.1523/jneurosci.42600-8.2009. Susa, K.J., Meissner, C.A., de Heer, H., 2010. Modeling the role of Rodin, 1987. Who is memorable to whom: a study of cognitive social-cognitive processes in the recognition of own- and disregard. Soc. Cogn. 5, 144–165. other-race faces. Soc. Cogn. 28, 519–533. Savaskan, E., Ehrhardtb, R., Schulzb, A., Walterc, M., Van IJzendoorn, M.H., Bakermans-Kranenburg, M.J., 2012. A sniff Schachingerb, H., 2008. Post-learning intranasal oxytocin of trust: meta-analysis of the effects of intranasal oxytocin modulates human memory for facial identity. J. Psyneuen administration on face recognition, trust to in-group, and 33, 368–374, http://dx.doi.org/10.1016/j.psyneuen.2007.12.004. trust to out-group. J. Psyneuen 37, 438–443, http://dx.doi.org/ Sheng, F., Liu, Y., Zhouc, B., Zhou, W., Han, S., 2013. Oxytocin 10.1016/j.psyneuen.2011.07.008. modulates the racial bias in neural responses to others’ Zak, P., 2012. The moral molecule. The source of love and suffering. Bio Psycho 92, 380–386. prosperity. Penguin Group, New York. brain research 1580 (2014) 22–56

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Oxytocin's role in anxiety: A critical appraisal

Kai MacDonaldn, David Feifel

University of San Diego, Department of Psychiatry, 140 Arbor Drive, CA 92103, USA article info abstract

Article history: A growing literature suggests that the oxytocin (OT) system may play a role in human Accepted 15 January 2014 anxiety states, anxiety-related traits, and moreover, that this system may be a target for Available online 24 January 2014 the development of novel anxiolytic treatments. However, studies of OT's acute and Keywords: chronic effects on various aspects of anxiety have produced mixed results. In this forward- “ ” Oxytocin looking review, we discuss the myriad phenomena to which the term anxiety is applied Anxiety in the OT literature and the problem this presents developing a coherent picture of OT's fi Attachment role in anxiety. We then survey several different elds of research that support the role of Evolution the OT system in human anxiety, including evolutionary perspectives, translational and Pharmacology neuroimaging research, genetic studies, and clinical trials of intranasal OT. As an “ ” Human outgrowth of this data, we propose a bowtie model of OT's role at the interface of social Animal model attachment and anxiety. We next direct attention to understudied brain regions and neural circuits which may be important to study in OT experiments in humans anxiety disorders. Finally, we conclude by proposing questions and priorities for studying both the clinical potential of OT in anxiety, as well as mechanisms that may underlie this potential. Crucially, these priorities include targeted proof-of-concept clinical trials of IN OT in certain anxiety disorders, including investigations of individual moderators of OT's anxiolytic effects (i.e. sex, genetic factors, and early experience). This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2014 Elsevier B.V. All rights reserved.

1. Introduction answer? Though a growing literature discusses the relation- ship between OT and anxiety—and frequently states that H.L. Mencken quipped that “for every complex problem there oxytocin is “anxiolytic”—the answer to this question is is an answer that is clear, simple, and wrong”. Human neither clear nor simple. As a case in point, research from anxiety disorders1 are complex problems. Might OT be an our own group has found both single-dose anxiogenic (Macdonald et al., 2013b) and chronic anxiolytic effects

nCorresponding author. (Feifel et al., 2011) of IN OT in humans with psychiatric E-mail address: [email protected] (K. MacDonald). disorders. In this forward-looking review, we survey several 1For the purposes of this review—and because most studies in literatures that inform the role of the OT system as a target humans have been done using the DSM IV system—unless otherwise specified, “anxiety disorders” refers to anxiety disor- ders in the DSM IV classification system. These include panic (footnote continued) disorder (PD), obsessive-compulsive disorder (OCD), social pho- (GAD), and post-traumatic stress disorder (PTSD). Recently, DSM 5 fi bia/social anxiety disorder (SOP), generalized anxiety disorder changed the classi cation of several of these disorders (notably, OCD and PTSD), added a diagnosis to the category (adult separa- http://dx.doi.org/10.1016/j.brainres.2014.01.025 tion anxiety disorder) and changed the diagnostic criteria for 0006-8993 & 2014 Elsevier B.V. All rights reserved. several disorders (see Table 1). brain research 1580 (2014) 22–56 23

for novel and discuss a variety of ways that the OT these terms are used to denote a wide variety of anxiety- system may be related to human anxiety-related traits and related phenomena in both animals and humans (Blume disorders. These bodies of research include: (1) evolutionary et al., 2008; Bosch, 2011; Grillon et al., 2013; Thompson biological perspectives on anxiety and the OT/social neuro- et al., 2011; Weisman et al., 2013a). Table 1 sets down a peptide system (including a bow-tie model of their overlap); number of these common anxiety-related terms and phe- (2) translational research on OT's relationship to mammalian nomena. As evidenced by the array of topics there, depending fear and anxiety; (3) studies of genetic variation in the OT on the knowledge domain and research field, anxiety can system; (4) neuroimaging studies of anxiety states, traits and refer to: (1) subjective/verbal statements (typically captured disorders; (5) human experiments using intranasal (IN) OT. in self-reports); (2) behaviors in nature or the lab; and/or (3) Building on these diverse literatures, we survey several brain biological and physiological measurements (i.e. elevated “regions of interest” found to be important across human cortisol, sympathetic arousal, amygdala activation) (Kagan, anxiety disorders, and propose them as future targets of 2008). Though these different domains all map onto a study of the anxiolytic effects of OT. We conclude by empha- triparate model of human anxiety (Lang and Davis, 2006), sizing that a key therapeutic question—whether chronic OT these shifting referents contribute to lack of clarity in the OT- or OT mimetics can give meaningful relief to humans with anxiety literature. Rather than proposing a solution to this anxiety and stress-related disorders—is yet to be answered. “tower of Babel” dilemma, we simply highlight the impor- Throughout this paper, though we emphasize the OT system, tance of clear, domain-specific language, and alert the reader we acknowledge the dynamic interplay between OT and AVP to this trend. systems (Neumann and Landgraf, 2012) (including the role of A second, related issue in the OT-anxiety literature is an AVP receptors in some of OT's anxiolytic effects (i.e. Mak often-mistaken perception that acute anxiolytic-type effects et al., 2012)). Among other reasons, this focus warranted by (i.e. short-term decreases in subjective fearful feelings or the fact that the majority of trials in humans that inform OTs acute neural effects—i.e. amygdala downregulation)—are anxiolytic effects have focused on the OT “half” of the social necessarily salutary, and the obverse: that acute neuropeptide system. anxiogenic-type effects are negative. As we have discussed in a prior publication (Macdonald et al., 2013a), the contrast 1.1. Overview: clinical anxiety disorders between the acute subjective anxiogenesis and amygdala activation seen with SSRIs (Burghardt and Bauer, 2013; Broadly speaking, anxiety disorders are the most prevalent McKie et al., 2005; Ravinder et al., 2013) and these agent's class of mental disorders, impose significant functional later-onset symptomatic benefits highlights the problems impairment, and typically have a chronic, recurrent course with this assumption. Coming from the other side of the (Kessler et al., 2005a, 2005b; Ravindran and Stein, 2010). argument, agents that have reliable short-term subjective Myriad factors influence their development, including genetic anxiolytic effects in humans (i.e. benzodiazepines) may and temperamental factors (Biederman et al., 2001; Ehringer impair fear learning (Stewart, 2005), and are not recom- et al., 2006; Fox et al., 2005; Hirshfeld-Becker et al., 2008; mended in the treatment of certain stress-related disorders Kagan et al., 2001; Norrholm and Ressler, 2009; Stein et al., (Tol et al., 2013). Therefore, when considering the potential of 2008). These factors interact in complex ways with early OT in the psychopharmacological treatment of human anxi- socioemotional experiences, resulting in the “programming” ety states and disorders, its acute effects should be disam- of hormonal systems like the HPA axis (Heim and Nemeroff biguated from potential long-term benefits after chronic (2001), Meaney et al. (2007), and Heim et al. (2008) for review) delivery. and the OT system (Champagne et al., 2001; Feldman et al., A third contributor to the mixed perception of OT's 2010; Macdonald, 2012; Pedersen and Boccia, 2002). Besides promise as an anxiolytic agent has to do with the hetero- the categorical anxiety disorders, anxiety symptoms are a geneous nature of the maladies classified as anxiety disor- common, transdiagnostic, functionally-impairing component ders (Craske et al., 2009). Though many of the DSM IV and of other many disorders including depression and schizo- DSM 5 anxiety disorders share phenomenological, psycholo- phrenia (D'Avanzato et al., 2013; Fava et al., 2008; Kessler gical, and neurobiological features (Craske et al., 2009; Etkin et al., 2005b). Current pharmacological treatments for anxiety and Wager, 2007; Menon, 2011; Mineka and Zinbarg, 2006; and anxiety disorders frequently have side effects or do not Sylvester et al., 2012), and though they are often treated with induce symptom remission, leading to a need for novel similar medication classes (i.e. SRIs) and therapies (i.e. antianxiety agents (Carson et al., 2004; Davidson et al., 2010). exposure-based therapies) (Craske et al., 2009; Ravindran and Stein, 2010), from multiple other perspectives (i.e. phe- 1.2. Defining terms: what is anxiety? what is an anxiety nomenology, factor analytical, genetics) they are a hetero- disorder? geneous group. Reflecting this fact, several conditions classified in the DSM IV as anxiety disorders have been Three linked topics contribute to a mixed perception of OT's reclassified in DSM 5 (see footnote above). Not only are there anxiolytic potential. The first issue—common in the neuros- significant within-category differences in the anxiety disor- ciences (Kagan, 2008)— is linguistic, and concerns vagaries in ders, but these afflictions share significant overlap with mood the use of the term “anxiety”. To wit, like the terms “proso- disorders (Craske et al., 2009; Kendler et al., 2011; Kendler cial” (Goodson, 2013), “trust” (Zak et al., 2004) and “love” et al., 2003; Krueger, 1999; Watson, 2005), a reality reflected in (Young, 2009), the terms "anxiety" (and therefore "anxiolytic") the NIH RDoC “negative valence systems” category (Insel are promiscuous and underspecified. In different contexts, et al., 2010). Notably, the RDoC criteria place OT in two 24 Table 1 – Varieties of anxious experience: terms used in the OT literature related to anxiety and stress-related disorders.

Construct/term Definition References related to OT Comments

“Anxiety” (and the Underspecified: depending on the See text 1. Neurobiologically, the activation or deactivation of some aspect related terms context, any of the below of the brain's central “fear system” (acute, phasic responses to “anxiolytic” and threat), or the related “anxiety system” (tonic, longer-term “anxiogenic”) responses to anticipation of danger) (Barlow, 2002; Davis et al., 2010) 2. In emotion literature, anxiety is often considered a secondary emotion in response to a primary emotional reaction (anticipatory response to fear or pain) (Barlow, 2002) 3. Components of human anxiety: Action tendencies (avoidance) À Appraisal/attention biases À Physiology (autonomic nervous system, HPA axis)

À 22–56 (2014) 1580 research brain Expression (facial, vocal) À Cognitions (worry) À Subjective feelings states (i.e. interoception of somatic À markers) (Barlow, 2002; Craske et al., 2009; Lang et al., 2000; Risbrough, 2010) Fear A phasic, primary, aversive state Guzman et al. (2013) and 1. In one influential model focusing on primary defensive states of of the nervous system Knobloch et al. (2012) the nervous system, fear is contrasted with anxiety, a tonic associated with threat to response associated with sustained fear and anticipation (Davis survival or pain et al., 2010). In human research, anxious anticipation is sometimes called dread (Berns et al., 2006; Carter and Krug, 2009; Mobbs et al., 2007) 2. The mammalian fear system is subserved by highly conserved subcortical brain networks (lateral and central amygdala, anterior and medial hypothalamus, PAG) (LeDoux, 1996; Panksepp, 1998) 3. Activation of this system coordinates and mobilizes facets of the organism—physiology (sympathetic nervous system, HPA axis), behavior, facial expression—to cope with and communicate immediate threat 4. Symptoms of clinical anxiety disorders are better detected in sustained rather than phasic fear paradigms (sustained fear) (Davis et al., 2010) Stress First defined as the non-specific Cardoso et al. (2013), Skopek et al. 1. Like “anxiety”, an underspecified term whose experimental response (“general adaptation (2012) definition has been perennially problematic (Koolhaas et al., syndrome”) of the body to noxious and Taylor et al., 2000 2011; Selye, 1936) stimuli (Selye, 1936) 2. Captured in the DSM 5 as part of “post-traumatic stress disorder” (American Psychiatric Association, 2013) 3. Related to disruption of homeostasis and the fight-or-flight response (Cannon, 1932) 4. Often recruits the neuroendocrine system, the autonomic nervous system and the (Koolhaas et al., 2011). Key elements of a restrictive definition of stressful situations include (1) unpredictability, (2) uncontrollability, and (3) environmental demands which exceed individual capacity (Koolhaas et al., 2011) Anxiety symptoms Subjective symptoms related to Lawson et al. (2013) 1. Can refer to symptoms contributing to a DSM diagnosis of anxiety or fear, typically anxiety disorder collected via self-report (i.e. STAI) 2. Single-question/Likert-type assessments of anxiety or fear (i.e. visual analog scales) do not tend to discriminate single-dose OT from placebo (MacDonald et al., 2011) 3. Dissociation between physiological/neurobiological signs of anxiety and subjective anxiety symptoms (response dyssynchrony) is common (see text for references and discussion) Endophenotypes of A restrictive definition states that de Oliveira et al. (2012a, 2012b), The definition of endophenotype varies in the anxiety literature anxiety disorders endophenotypes should be laboratory-based Labuschagne et al. (2010, 2011) and from more to less restrictive, with the less restrictive types biomarkers that are (1) quantifiable; (2) associated with Panksepp (2006) sometimes called intermediate phenotypes (see footnote in text) genetic risk for the disorder; (3) abnormal in patients and probands and (4) relatively invariant (Bearden and Freimer, 2006)

Anxiety-related A variety of behavioral responses linked Figueira et al. (2008), Grillon et al. 1. Correlation of animal or human behavior with (1) subjective 22–56 (2014) 1580 research brain behaviors/anxiety- to anxiety disorders and used in animal (2013), Mantella et al. (2003) and anxious feelings in humans; (2) human anxiety disorders; or (3) like behavior and human models (i.e. fear-potentiated Slattery and Neumann (2010) pharmacological agents used to treat human anxiety is imperfect startle (FPS)) (see text) (see discussion in Baas et al. (2009) and Fernando and Robbins (2011)) 2. The effects of OT on FPS in humans have been examined in several human studies (Acheson et al., 2013; Grillon et al., 2013). The FPS model has both advantages and disadvantages in terms of screening for anxiolytics (Grillon et al., 2009; Klumpers et al., 2010; Baas et al., 2009) (and see text) 3. The effects of OT have been examined in many animal models of anxiety (Lin, 2012; Rotzinger et al., 2010) Separation anxietya “overconcern about their offspring and Costa et al. (2009) 1. Though it has unclear homology to adult separation anxiety, spouses… experience marked discomfort animal models use maternal separation—which causes changes when separated from them” in OT systems (Kojima et al., 2012; Lukas et al., 2010)—to model early life stress (Zhang et al., 2012) 2. Adult separation anxiety disorder (ASAD) is not uncommon (Bogels et al., 2013), and is now listed in DSM 5 anxiety disorders (American Psychiatric Association, 2013) 3. Classification of behaviorally inhibited children at 2 years is composed of clinging to and remaining proximal to the mother and approach of novel stimuli (Schwartz et al., 2012) Anxiety disorder One of the anxiety disorders Feifel et al. (2011), Guastella et al. The DSM 5 recently reclassified several anxiety disorders (see codified in the DSM system (2008), Labuschagne et al. (2011) and footnote in text) Macdonald and Feifel (2013) for review Anxiety-related See text for discussion temperaments Anxious A trait-like tendency to avoid Poirier et al. (2013) More than other anxiety-related temperaments, anxious temperament (aka novel or unfamiliar situations, temperament is related to behavioral observations, and therefore “inhibited and demonstrate increased physiological can be used in translational research in animals and very young temperament”) responsiveness and behavioral inhibition (preverbal) humans (i.e. Oler et al., 2010) to mild threat Attachment anxiety 25 26 brain research 1580 (2014) 22–56

domains: negative valence systems and social processes (Insel et al., 2010), creating a shared space of interest in terms of the role of OT in human anxiety (Figs. 1 and 2). In the Ball et al., and future, human studies will likely use a combination of DSM 5 ” and RDoC-influenced criteria and endpoints (Insel et al., 2010), a dual agenda which will create both challenges and opportunities. Ultimately, though the abovementioned issues are impor- tant, the largest and most glaring impediment to under- harm avoidance “ standing the role of OT as a bona-fide treatment for human anxiety disorders is lack of direct clinical research. In fact, a

), and a number of parameters of search for published studies treating patients with anxiety ) describing trait-like tendency toward

” disorders with chronic (multiple-week) OT yields a total of only 26 patients (Table 2). Almost half of these are negative

ies relating to that aspect of anxiety. See text for studies of patients with OCD (den Boer and Westenberg, 1992; Epperson et al., 1996), a disorder now considered have significantly different neurobiological substrates than other anxiety disorders (Brennan et al., 2013; Saxena and Rauch, Wachleski et al., 2008 ; 2000). As such, though the studies reviewed herein suggest anxious temperament Along with attachment avoidance,parameters one demonstrated of the in two some attachment individual studies to responses moderate to certain oxytocin (see text) neurobiology (see text) One of a cluster of“ terms (see also tension, anxiety, and negative affectivity Harm avoidance is associated2002 with anxiety disorders ( that the OT system may have a role in the treatment of anxiety disorders, with the exception of a few pilot trials (Feifel et al., 2011), studies that directly address key clinical

Love questions about the role of OT in human anxiety disorders , are still sorely lacking. Wang et al.

Weisman et al. 1.3. Evolution, anxiety, and social neuropeptides and and

Kiss et al. (2011) The long evolutionary history of fear, anxiety, and the social , Weisman et al. (2013a) nonapeptides (OT and AVP) strongly inform OT's potential as and an anxiolytic agent. From this evolutionary and phylogenic perspective, it has long been noted that many human anxiety symptoms and disorders have their roots in highly conserved mammalian “defense systems” (Belzung and Philippot, 2007; Bartz et al. (2011) et al. (2012) Stuebe et al. (2013) (2013a) Stankova et al. (2012) (2013) Deakin and Graeff, 1991; Lang and Davis, 2006; Panksepp, 1998). The most ancient roots of these brain-based systems have both appraisal (i.e. low-level, reflexive valancing (Robinson, 1998)) and hormonal aspects which—like OT homologs—can be found in simple invertebrates (Beets

), et al., 2013; Belzung and Philippot, 2007; Panksepp, 1998; Yamashita and Kitano, 2013). Over millennia, these primitive ) mechanisms become intercalated in vertebrates with homeo- static mechanisms (i.e. the autonomic nervous systems), and in mammals, aspects of these systems are exapted, modified, Cloninger, 1987 and repurposed into mechanisms for social signaling, social perception, attachment, and pairbonding (Chang et al., 2013; Emery, 2000; Porges, 2003b; Shultz and Dunbar, 2007; Brennan et al., 1998 Tomasello et al., 2007). Many of these defense and safety- related mechanisms (i.e. eye contact, vocalizations, facial expressions) have been shown to be correlated with or

nition References related to OTmodi Comments fied by OT (Domes et al., 2013b; Seltzer et al., 2010; fi

A trait measure of sensitivityworry about to or rejection or losswith intimates of relationships ( assessed with Cloninger's temperament and character inventory ( and associated with excessivecautiousness, worrying, pessimism, shyness and fearfulness Shahrestani et al., 2013). A critical aspect of the evolutionary advance in placental mammals was the comingling of brain nition. fi reward mechanisms with chemistries related to sex, birth and nursing (including OT-related systems (Dolen et al., ) 2013)); these allowed infants and mothers to sustain dyadic behavioral “addiction” (Insel, 2003) and “symbiotic regula-

continued tion” (Hofer, 1994b; Porges ad Carter, 2012). Importantly, development of these social attachments necessitated the Based on DSM 5 de Trait anxiety Measured with STAI, trait version A variety of terms are used in the literature pertaining to OT's role in human fear, anxiety, and anxiety disorders. References in table refer to OT stud discussion. a Construct/term De Harm avoidance A heritable temperamental trait modulation of older defense-related systems: deactivation of Table 1 ( amygdala (Moriceau and Sullivan, 2005), regulation of brain research 1580 (2014) 22–56 27

Fig. 1 – As captured in the RDoC criteria, different aspects of the larger field of OT-related research relate to both social/ attachment functions, as well as anxiety and anxiety disorders. nThough attachment-related topics and social processes and behaviors are not synonymous, they are combined here for simplicity.

sympathetic tone via myelenated vagus (Porges and Carter, Chang et al., 2013; Panksepp and Northoff, 2009; Shultz and 2012), and relational entrainment of the HPA axis (Feldman, Dunbar, 2007; Tomasello et al., 2005). These cortical struc- 2012; Hofer, 1994a; Insel, 2003). Here, in ancient brain reward tures play a central role in our current understanding of and regulatory systems, we find the neural roots of sustained human anxiety disorders (see Section 4.7), though our under- attachment bonds, as well as the microcircuitry wherein standing of the role of OT in these areas is nascent (but see social attachments function as ‘hidden’ regulators of the recent, seminal study of cortical effects of OT (Zheng, 2014)). physiology and behavior of offspring (Feldman, 2012; Hofer, To summarize: tracing the phylogenic pathway of these two 1994a; Porges and Carter, 2012). highly-conserved brain systems –defense and attachment— From an evolutionary perspective, then, maternal-infant alongside the remarkably conserved OT system highlights sev- attachment became a new “safety signal”, creating links eral pathways through which OT may have effects on the between single-skull survival-oriented defense systems and expression, phenomenology and development of human anxi- OT-associated attachment systems (Hofer, 1994a; Panksepp eties at different levels of the evolved neural axis (Bethlehem et al., 1997; Porges, 2003a). This evolutionary fact introduces et al., 2013; Kirsch et al., 2005; Riem et al., 2013; Zink and Meyer- an interesting dialectic into OT research and social neu- Lindenberg, 2012). This idea is further developed in Section 3,in roscience: though for the developing infant attachment is a the context of a “bowtie” model of OT. survival necessity, outside of this bond, and throughout the rest of mammalian life, conspecifics and other minded beings 1.4. Causal thickets: OT and levels of analysis are a more ambiguous social signal, a potential source of reward and pleasure as well as competition, stress, and Alongside the “long view” provided by a phylogenic and danger. This dialectic informs the potential “anxiolytic” evolutionary perspective on OT's role in fear and anxiety, effects of OT vis a vis its ability to increase the salience of one can place a more present-oriented perspective on different social stimuli (Groppe et al., 2013). That is, increasing the “levels of analysis” present in the OT-anxiety literature salience of social stimuli in different contexts may have very (Table 3). In term of the broader goal of understanding OT's different survival implications, as well as different implica- role in the development and treatment of anxiety disorders tions for fear and anxiety. within a human lifespan, it is worth noting that the more we A finally step in the evolutionary process discussed above understand the details of a system, the greater the number of came with the evolution of integrative cortical brain struc- potential cross-level interactions, emergent system properties tures supporting second and third-level representations of and feedback loops are created (Kendler, 2005, 2008). These body state, higher levels of consciousness, and more complex interactions create recursive etiological pathways advanced social cognition. These include structures and and complex “causal thickets” (Kendler, 2005, 2008). In the “modules” that instantiate aspects of self-awareness, aware- case of OT, these multi-level dynamics may be the best ness of the minds of others, and future-oriented cognition theoretical model with which to understand emerging studies (Belzung and Philippot, 2007; Brune and Brune-Cohrs, 2006; of OT's role in gene x environment interactions (for example, 28 brain research 1580 (2014) 22–56 Diverse socially

SOCIAL/ATTACHMENT REWARD NEUROCHEMISTRIES t

a (SOCIAL AVERSION/SOCIAL SALIENCE/BIOBEHAVIORAL SYNCHRONY) v i t c a i l u m i t - s Orgasm

1. relevant reactions regulated by knot l Dopamine Opiates 1. Social appraisal/salience

a 2. Birth i 2. Facial attention (eye region) c 3 Nursing Gonadal Steroids o 3. Social reward

s 4. Mother-Infant bonding / 4. Alteration of social- d 5. Father-infant bonding HYPOTHALAMIC

e memory networks t 6. Pair-bonding NEUROPEPTIDE SYSTEM

a 5. Autonomic nervous system l

HPA Axis (I.E. OT, AVP) e 7. In-group social connections regulation the knot r 8. Touch 6. Social approach/avoid 9. Generosity behavior 7. Pain /distress regulation Serotonin Glutamate Diverse attachment-Diverse e Diverse brain regions mediating fear / anxiety DEFENSE/FEAR- RELATED NEUROCHEMISTRIES

1. PAG*

2. Other Brainstem Metabolic/cardiorespiratory 1. Dopamine Opiates regions** stress(i.e. high sodium) 3. Dorsal Raphe 2. Loss of attachment bond

Norepi 4. Locus Ceruleus 3. Social threat/evaluation HYPOTHALAMIC NEUROPEPTIDE SYSTEM 5. Amygdala* 4. Pain (physical or social)

HPA Axis HPA (I.E. OT, AVP) 6. BNST 5. Predator threat 7. Hippocampus 6. Threat to mating/attachment pair Serotonin GABA 8. Insula*

9. Prefrontal cortex Diverse stimuli and conditions that signal danger signal that conditions and stimuli Diverse Shared Brain Regions and Processes

Social & Social Attachment processing Stimuli HYPOTHALAMIC & Behavior NEUROPEPTIDE SYSTEM (I.E. OT, AVP) Brain Danger Regions Stimuli Shared Social/Danger Signals

Fig. 2 – The central OT/social neuropeptide system is embedded in a layered, interactive central network, and is an example of the knot in a bowtie architectural structure. Bowtie architectures are acharacteristicofrobust,evolvablesystemsinbiology(i.e.the social attachment system (a) and fear/defense system (b)), and have extensive ‘fan-in’ inputs and a diverse array of ‘fan-out’ outputs. Left and right sides of the bowtie are connected by a highly-conserved core network whose components have modular, weak linkages (double arrows). Notably, information does not flow solely from left to right: extensive local and global bidirectional influences (i.e. genetic and epigenetic differences within the OT system, hormone levels, and receptor densities) ultimately regulate its dynamics. The shared knot (c) between these evolved and highly conserved systems and the overlap between left and right sides illuminates how OT systems may play a role in human fear and anxiety. Overlapping areas on left of (c) are shared safety/social signals; overlapping areas on right are brain regions and reactions with relevance to both social processing and anxiety disorders. See text for discussion and references. n Brain regions also associated with social processing; nnSee Fig. 3 for details. BNST bed ¼ nucleus of the stria terminalis; GABA gamma-aminobutyric acid; HPA hypothalamic-pituitary axis; NE norepinephrine, ¼ ¼ ¼ PAG periaqueductal gray. Figure reprinted and modified with permission from Csete and Doyle (2004). ¼ brain research 1580 (2014) 22–56 29

Table 2 – Human treatment studies of relevance to IN oxytocin's effects on anxiety.

Reference(s) Population N, sex, age Parameter Dosing Findings (ifo18)

Normals Heinrichs 37M Subjective anxiety, CT 24 IU After TSST, OT group showed lower et al. (2003)c response (TSST) CT and higher calmness ratings, similar to social support Kirsch et al. 15M Brain responses to 27 IU IN OT reduced connectivity (2005)b fearful or threatening between amygdala and brain stem visual images regions involved in autonomic functions Petrovic et al. 27M Evaluative 32 IU Post-acquisition OT diminished (2008) conditioning conditioned negative subjective responses to faces, responses and attenuated neural activity amygdala, anterior cingulate activation relative to PBO Ditzen et al. 47M, 47F Behavioral and 40 IU OT significantly improved negative (2009) physiological to positive communication ratio responses to conflict and salivary CT levels discussions Quirin et al. 36M Stress-induced (TSST) 24 IU Individuals with lower ERA (based (2011) CT increases based on on Action Control Scale) showed (ERA) blunted CT response after OT de Oliveira 28M Subjective anxiety, 24 IU OT caused decreased anticipatory et al. (2012b) skin conductance, (pre-test) subjective anxiety and heart rate (SBP, DBP) decreased skin conductance prior to a validated model of public speaking test called the SPST (McNair et al., 1982) (see text for details) de Oliveira 45M Subjective anxiety, 24 IU OT prevented increase in subjective et al. (2012a) DBP, SBP, CT anxiety induced by CO2. See text for

notes on CO2 model (text) Linnen et al. 48F, 48M Affective and cortisol 24 IU Task did not elevate CT, and IN OT (2012) response to the YIPS did not impact mood. IN OT was task associated with decreased CT over course of experiment (see text for discussion) Sripada et al. 15M Resting-state fMRI 24 IU IN OT was found to significantly (2012) increase resting state connectivity between the amygdala and specific regions of the frontal cortex in normal subjects, while having only negligible effects on coupling with other brain regions Rupp et al. 59F (29 Amygdala sensitivity 24 IU Both amygdala activation and (2014) postpartum, to negative visual subjective negative arousal 30 stimuli (IAPS pictures). demonstrated reductions in nulliparous) Urinary cortisol levels postpartum vs. nulliparous women. Postpartum women had lower stressed CT levels. IN OT eliminated the parturitional difference in stress response Cardoso et al. 17M Exercise-induced CT 24 IU, 48 IU Exercise-induced CT increase (2013) attenuated by 24 IU but not 48 IU dose Weisman 35M (fathers) Fathers and infants 24 IU IN OT increased paternal HPA et al. (2013a) CT response to social responses to social stressor (still- stressor face paradigm), and altered infant's CT response to still face stressor. Effects on infants moderated by quality of attachment relationship Grillon et al. 19F, 24M (FPS) 24 IU Compared with PBO and AVP, OT (2013) increased FPS to unpredictable (but not predictable) shock Acheson et al. 21F, 23M 24 IU OT increased extinction recall, but (2013) did not facilitate extinction. OT had 30 brain research 1580 (2014) 22–56

Table 2 (continued )

Reference(s) Population N, sex, age Parameter Dosing Findings (ifo18)

Fear conditioning and no effects on self-reported anxiety fear extinction in FPS or expectancy task

Single-dose trials in patients with brain-based illness Bartz et al. Borderline 10F, 4M Neuroeconomic trust 40 IU OT impeded trust and prosocial (2011) personality game behavior, moderated by attachment disorder anxiety and avoidance Simeon et al. Borderline 6F, 8M Post stressor 40 IU OT attenuated of subjective post- (2011) personality subjective mood, CT stressor dysphoria, and caused a disorder response trend to decreased CT. Results moderated by trauma, self-esteem, and attachment style Bertsch et al. Borderline 40F Eye movements and 26 IU OT normalized several disorder- (2013) personality fMRI responses to specific responses to angry faces, disorder emotional faces causing a reduction in posterior amygdala hyperactivation, as well as a reduced attentional (visual) bias toward the eye region of angry faces Hall et al., Fragile-X 8M, age 13– Eye gaze frequency, 24–48 IU Eye-gaze improved with 24 IU; CT Reiss (2012) syndrome 28 HR, HR variability, CT decreased with 48 IU Pincus et al. Major 8F Reaction time and 40 IU Compared with controls, depressed (2010) depressive brain responses (fMRI) patients doing the RMET activated disorder to RMET higher order cognitive areas and insula with OT MacDonald Major 17M Subjective anxiety 40 IU OT caused an increase in STAI over et al. (2013b) depressive (STAI) the course of a 20 min therapy disorder session McRae-Clark Marijuana 12M, 4F Stress responses, 40 IU OT administered prior to TSST et al. (2013) dependence subjective anxiety, resulted in decreased MJ craving craving, CT, DHEA scores. Lower anxiety response to levels TSST was observed, but subjective stress ratings not reduced. DHEA levels decreased with IN OT Koch et al. Post-traumatic 21M Amygdala reaction to 40 IU OT reduced right amygdala (2013) stress disorder fearful and angry activation during fearful and angry faces faces Pitman et al. Post-traumatic 43M Physiologic responses 20 IU OT subjects had the lowest mean (1993) stress disorder (HR, GSR, facial EMG) physiologic responses to personal to personal trauma combat imagery prompts, verses prompts placebo and IN AVP-treated subjects Labuschagne Social anxiety 18M Brain responses (fMRI) 24 IU Patients exhibited bilateral et al. (2010) disorder to emotional face amygdala hyperactivity to fearful (generalized) matching task with faces; OT normalized this effect fearful, angry, happy faces Labuschagne Social anxiety 18M Brain responses (fMRI) 24 IU Patients had heightened activity to et al. (2011) disorder to emotional face sad faces in medial prefrontal (generalized) matching task of cortex and anterior cingulate happy and sad (vs. cortex; OT reduced this neutral) faces hyperactivity Guastella et al. Social anxiety 25M Self-rated aspects of 24 IU OT-treated subjects demonstrated (2009) disorder social anxiety, speech improved self-evaluation of performance and appearance and speech appearance performance; these benefits did not generalize into a sustained positive effect over exposure therapy alone

Multiple-dose trials/case reports Pedersen et al. Alcohol 9M 2F Alcohol withdrawal 24 Twice- OT-treated patients required less (2013) dependence scores, lorazepam use daily for 3 lorazepam, had lower alcohol days brain research 1580 (2014) 22–56 31

Table 2 (continued )

Reference(s) Population N, sex, age Parameter Dosing Findings (ifo18)

withdrawal scores, and lower subjective distress Feifel et al. Generalized 7M, 6F HAM-A 20 Twice- Males showed significant decrease (2011) anxiety disorder daily for 1 in anxiety at week 2, females week, then showed trend increase in anxiety, 40 IU twice- with trend significance drug x daily for 2 gender interaction weeks Scantamburlo Major 1M HAM-D, STAI, Q-LES-Q Up to 36 IU Adjunctive OT improved depressive et al. (2011)a depressive over several and anxiety symptoms and quality disorder weeks of life over the course of weeks den Boer and Obsessive 3M, 9F Obsessions and (1) 18 IU IN for 6 Westenberg compulsive compulsions weeks (dosed (1992) disorder four times daily) (2) 2 M treated with 54 IU No effect on symptoms Epperson et al. Obsessive 3F, 4M Obsessive compulsive 160 IU or No change in obsessive-compulsive (1996) compulsive disorder symptoms, 320 IU IN disorder symptoms. OT subjects disorder anxiety, mood and (divided four had a statistically significant memory times daily) improvement in BDI for one week Macdonald Social anxiety 1M Social anxiety 20 IU twice Improvement in several areas of and Feifel disorder symptoms, sexual daily over sexual function, though no benefit (2012)a function several weeks in social avoidance or anxiety Epperson et al. Trichotillomania 2F Trichotillomania 160 IU No difference in trichotillomania (1996)a symptoms (divided four symptoms times daily) for one week

Abbreviations: AVP arginine vasopressin, BDI Beck Depression Inventory, CO Carbon Dioxide, CT cortisol, DBP Diastolic Blood Pressure, ¼ ¼ 2 ¼ ¼ ¼ DHEA Dehydroepiandrosterone, EMG electromyography, ERA Emotion Regulation Abilities, F female, fMRI functional Magnetic Reso- ¼ ¼ ¼ ¼ ¼ nance Imagery, FPS Fear-Potentiated Startle, GSR galvanic skin response, HAM-A Hamilton Rating Scale for Anxiety, HAM-D Hamilton ¼ ¼ ¼ ¼ Rating Scale for Depression, HR heart rate, IN intranasal, IU international units, M male, OT oxytocin, Q-LES-Q Quality of Life ¼ ¼ ¼ ¼ ¼ ¼ Enjoyment and Satisfaction questionnaire, RMET Reading the Mind in the Eyes Test, SBP Systolic Blood Pressure, STAI State-Trait Anxiety ¼ ¼ ¼ Inventory, YIPS Yale International Stressor, TSST Trier Social Stress Test. ¼ ¼ a Case reports. b Given the large number of neuroimaging studies showing OT effects on amygdala and other anxiety-relevant structures, only selected recent studies of these effects are included. Readers are referred to recent reviews for a complete list of relevant studies indicating these effects (Bethlehem et al., 2013; Zink and Meyer-Lindenberg, 2012). c A large number of studies have examined IN OT effects on cortisol and the HPA axis in both normal populations and a wide variety of psychiatric populations (Burri et al., 2008; Cardoso et al., 2013; de Oliveira et al. (2012b, 2012a); Ditzen et al., 2009; Heinrichs et al., 2003; Linnen et al., 2012; Quirin et al., 2011; Simeon et al., 2011). For reasons of space, not all are listed here. See text for discussion. see Brune (2012) and Nolte et al. (2011)). Examples of OT- levels and over multiple timeframes—from evolutionary (mil- related studies exploring between-level causal loops germane lennial) to developmental (individual lifespan) to dynamic to human anxiety are: (1) translational studies examining the (seconds to milliseconds)—informs our understanding of the effects of OT knockout on maternal behavior (Macbeth et al., microarchitecture and complex, multi-level interactions that 2010; Pedersen et al., 2006); (2) experiments examining the may underlie different aspects of human anxiety and the OT epigenetic impact of maternal behavior on the development of system's role in it (Table 3). stress systems related to anxiety (Champagne, 2008); (3) findings relating aspects of OT function to maternal depres- sion which then influences children's later development of 2. Translational animal research on OT and anxiety disorders (Apter-Levy et al., 2013; Jonas et al., 2013; anxiety Skrundz et al., 2011); (4) studies of the impact of early relational trauma on neural hubs (i.e. amygdala) where OT For reasons of space, we will not comprehensively review the has prominent effects ((Dannlowski et al., 2012; Olsavsky et al., large, diverse body of translational and basic science research 2013), and see Nolte et al. (2011) and references therein). Each that informs the role of the OT system in anxiety (readers are of these experiments explores different facets of the question directed to Lin (2012), Neumann and Landgraf (2012), “what is the role of the OT system in human anxiety?” In Rotzinger et al. (2010) and Stoop (2012)). In one recent review summary, examining aspects of the OT system at multiple summarizing a variety of animal models of both trait-like 32 brain research 1580 (2014) 22–56

Table 3 – Oxytocin and anxiety: recursive interactions and levels of analysis.

Level of analysis Factors relevant to OT and anxiety

1. Molecules/receptors 1a. OT, OT mimetics (Ring et al., 2010), OT releasers (Bagdy and Kalogeras, 1993), OTR gene variants (Kumsta et al., 2013; Walum et al., 2008) 1b. AVP receptor-mediated effects (Hicks et al., 2012; Mak et al., 2012; Poirier et al., 2013; Schorscher-Petcu et al., 2010; Zheng et al., 2013) 1c. Other aspects of OT system function (i.e. CD38 (Jin et al., 2007), second messenger systems (Okimoto et al., 2012; van den Burg and Neumann, 2011)) 1d. Regional distribution of OT receptors (Boccia et al., 2013; Liu et al., 2005; Zheng et al., 2013), different timescales of OT effects (wired vs. diffusion-mediated effects) (Ludwig and Leng, 2006; Stoop, 2012) 2. Cells 2a. OT neurons (Boccia et al., 2013; Huber et al., 2005; Knobloch et al., 2012; Viviani et al., 2011) and their regulation by sex hormones, monoamines, etc. (Brown et al., 2013) 2b. OT effects on CRF neurons (Dabrowska et al., 2011; Dabrowska et al., 2013) 2c. OT effects on 5HT neurons (Dolen et al., 2013; Eaton et al., 2012; Yoshida et al., 2009) 2d. OT effects on interneurons (Owen et al., 2013) 2e. OT effects on neurogenesis (Leuner et al., 2012). 3. Nodes and networks (within-brain dynamics) 3a. Functional impact of IN OT on evoked activity in amygdala, insula, mPFC (Koch et al., 2013; Labuschagne et al., 2010, 2011) 3b. Effects of OT on evoked amygdala-brainstem (Kirsch et al., 2005), amygdala-PFC (Sripada et al., 2012), and amygdala-insula (Riem et al., 2011) connectivity 3c. Effects of OT on resting state connectivity (DMN) (Sripada et al., 2012) and other functional networks 3d. Effects of genetic variation in OT systems (i.e. OTR) on connectivity, structure, function (Wang et al., 2013) 3e. Role of epigenetic changes in OT system on brain regions and network structure and function: canalization, plasticity, resilience (Brune, 2012; Jack et al., 2012; Kumsta et al., 2013; Unternaehrer et al., 2012; Veenema, 2011) 4. Neurobiology of attachment relationships 4a. Studies of OT and maternal (Ring Atzil et al., 2011; Strathearn et al., 2009), paternal (dynamics between bonded brains) (Kenkel et al., 2013; Weisman et al., 2013b) and infant behavior 4b. Effects of parental and infant OT genetic variation on future anxiety-related states (Rodrigues et al., 2009) 4c. Effects of early interpersonal trauma on anxiety-related networks (Burghy et al., 2012; Grant et al., 2011; White et al., 2012) and Nolte et al. (2011) for review 4d. Adult pair-bonding (Schneiderman et al., 2013, 2012) and social loss (Norman et al., 2010) 5. Neurobiology of social relationships (neural 5a. Effects of social support and social isolation (Grippo et al., 2012, 2009; Kim et al., dynamics related to wider social networks) 2010; Pournajafi-Nazarloo et al., 2013) 5b. In-group/out-group (De Dreu et al., 2011) and ethnic/racial effects (Inoue et al., 2010; Tost et al., 2011) on anxiety disorders 5c. Impact of social hierarchies/social defeat on anxiety disorders (Guzman et al., 2013; Lukas et al., 2011)

Data from multiple levels of analysis of the OT system inform our understanding of its role in the development and treatment of human anxiety states and disorders. Experiments informing these levels occur in a variety of different populations and models, including: (1) translational research in a variety of species; (2) anxiety/stress states in normal humans; (3) anxiety-related traits in non-clinical populations; (4) anxiety disorders in humans; (5) other disorders with anxiety as a component (i.e. borderline personality disorders, eating disorders). Importantly for an understanding of the role of the OT system in human , bidirectional, between- and within-level effects create complex, recursive causal loops (Kendler, 2008). See text for discussion

harm avoidance and conditioned fear extinction, a pattern of germane to the topic of human anxiety disorders. First, promising (but mixed) acute and chronic anxiolytic results animal research with OT is invaluable in our understanding emerged, overall supportive of future studies in humans of the OT system dynamics as it allows the disambiguation of (Rotzinger et al., 2010). Since this particular review, much separate neuropeptide receptor effects (through the use of important animal research informing OT's role as an anxio- directed application of OT agonists and selective vasopressin lytic has been done (Ayers et al., 2011; Guzman et al., 2013; agonists (Mak et al., 2012; Smith and Wang, 2013)), the use of Knobloch et al., 2012; Okimoto et al., 2012; Viviani et al., 2011) specificknockoutstrainsofanimals(Amico et al., 2004; Mak and (Neumann and Landgraf (2012) for review). et al., 2012; Mantella et al., 2003), as well as other novel Though space does not permit a comprehensive review of techniques unavailable in humans (i.e. optogenetics (Knobloch the details of this vital body of preclinical work, we highlight et al., 2012), gene deletion (Dolen et al., 2013), variations in three key issues that emerge from translational research maternal care (Champagne et al., 2001)). Secondly, though brain research 1580 (2014) 22–56 33

mammalian defense and neuropeptide systems and circuitries to-one’ inputs, and fan-out, ‘one-to-many’ outputs, joined by a have significant commonalities, there are also significant intras- central ‘knot’ made of universal building blocks or common pecies differences in OT-related parameters including aspects of currencies (here, the social neuropeptide systems OT and AVP sociality (Goodson, 2013)andOTreceptordensity(Ophir et al., (Csete and Doyle, 2004; Kitano, 2004; Tieri et al., 2010))(Fig. 2). It 2012; Rosen et al., 2008; Ross et al., 2009; Y. Wang et al., 2013; has been suggested that this bowtie pattern evolved to optimize Young and Wang, 2004). This point suggests cautious extension survival in highly variable environments, constraining and of experimental findings between species. Third, translational organizing fluctuating inputs and flexible outputs through a research has uniquely informed our understanding of the relatively small numberofcentralelements(Csete and Doyle, difference between short-term and long-term OT treatment, as 2004; Kitano, 2004; Tieri et al., 2010). Bowtie architectures have well as the potential longer-term impact of OT treatment during been used to describe a variety of different biological systems, critical developmental windows. For example, in an evocative including metabolism (M. Csete and Doyle, 2004), the immune study with alcohol-preferring rats, OT given during a key devel- system (Tieri et al., 2010)andtheinterfaceofobesityand opmental epoch altered behavior (beer intake), anxiety, and addiction (Volkow et al., 2013). Here, we propose that the sociability later in life (Bowen et al., 2011). Studies using viral conceptual model of a bowtie architecture can be helpful in vectors to increase OTR density in the NAcc have demonstrated organizing and understanding interactions between OT/AVP dissociation between acute effects of this intervention (increased systems and the multi-component, multi-level neurobiological partner preference) and chronic effects (modulation of allopar- systems that subserve human fear and anxiety. As mentioned ental behavior) (Keebaugh and Young, 2011). A final example in above, for simplicity we focus here on OT, though a balance this arena are several recent, cautionary studies in which acute between OT and AVP is an vital concept within the social IN OT caused increases in the social behavior of male voles, neuropeptide knot (Neumann and Landgraf, 2012), and many whereas long-term treatment resulted in behavioral deficits of OT's central actions may be mediated through its binding to (Bales et al., 2013; Huang et al., 2013)andreductioninbrain AVP1a receptors (Mak et al., 2012). OTRs (Huang et al., 2013). All told, then, though translational Regarding the characteristics of the central node of this research has generated a promising signal regarding OT's architecture, the core of a bowtie is comprised of a densely anxiolytic potential, it has also raised several important concerns connected, small-world network, which includes a variety of regarding long-term consequences and individual moderators of related, evolved control systems (surrounding arrows, Fig. 2) OT's effects. which have weak linkages with the core process (here, the OT/ AVP system) (Brown et al., 2013; Kitano, 2004). These surrounding processes often perform similar roles (i.e. DA and OT both 3. OT, attachment, and anxiety: bow ties that signaling salience), therewith creating redundancy or degener- bind? acy: different biological elements demonstrating functional homology. This principle makes the system resistant to single- As mentioned in Section 1.3,OT,fearandanxietyareyokedby component failure and allows positive and negative feedback to the fact that over long spans of evolutionary time, neural amplify signal-to-noise ratios (Kitano, 2004; Tieri et al., 2010). In systems with both unique and shared elements evolved to the case of OT, critical ‘within-node’ interactions include regulate both mammalian social reward/attachment2 as well dynamic relationships with CRF, AVP, opiates, serotonin, dopa- as defense/fear. Both of these systems demonstrate a funda- mine, GABA, glutamate, and gonadal steroids (estrogen and mental feature of complex, evolvable systems: robustness testosterone) ((Brown et al., 2013)).3 Multi-level, versatile, (Belzung and Philippot, 2007; Deakin and Graeff, 1991; Porges, 1998; Young, 2002). Defined as the cross-species preservation of 3Anatomically, the paraventricular nucleus (PVN) of the particular characteristics despite differences in specificcompo- hypothalamus is literally a “knot” in a neurobiological “bowtie”: nents or the environment (Csete and Doyle, 2002; Kitano, 2004), it receives direct and indirect input from a vast number of brain regions and exerts wide influence throughout the CNS by coordi- robustness abets translational OT research. Both within and nating sensory, neuroendocrine, and behavioral responses at outside of biology, robust, evolved systems exhibit an architec- multiple levels of the neural axis (Argiolas and Melis, 2013; tural framework called a bowtie, which features fan-in, ‘many- Swanson, 2000; Swanson and Sawchenko, 1983). Within the PVN are two populations of OT neurons: magnocellular secretory 2In the social neuropeptide literature, different terms are used neurons (MSNs), and parvocellular neurons of two types: pre- for the suite of shared neural circuits that underlie cross-species autonomic and neuroendocrine (Luther et al., 2002; Swanson and social and attachment-related phenomena. These include: (1) the Sawchenko, 1983). Details of the dynamic regulation of MSNs are “social brain” (Dunbar, 2003); (2) the “social reward/attachment beyond the scope of this review, (see Brown et al., 2013 for figure, system”, which undergirds maternal-infant and mating-pair references and a comprehensive review). Centrally-projecting bonds (Bartels and Zeki, 2004; Bowlby, 1969/1982; Feldman, parvocellular OT neurons coordinate information from the PVN 2012; Insel, 1992; Nelson and Panksepp, 1998; Panksepp et al., to its targets via “wired”, direct neurotransmission to areas like 1978); and (3) the “social behavior network”, a collection of limbic the brainstem and spinal cord, and receive both adrenergic and forebrain regions which mediate a wider range of social phenom- noradrenergic innervation from the brainstem, bed nucleus of the ena (i.e. sexual, parental, aggressive behavior) (Newman, 1999). stria terminalis, and adjacent hypothalamic areas (Ludwig and Due to the significant amount of translational research on OT’s Leng, 2006; Stoop, 2012; Swanson and Sawchenko, 1980). Regula- role in the mammalian attachment/social reward network tion of the parvocellular arm of the central OT system has been (Feldman, 2012; Insel, 1997; Panksepp et al., 1997; Pedersen and less well-studied than regulation of MSNs, though extant evi- Prange, 1979), we here refer specifically to the subcortical net- dence indicates a complex, dynamic regulatory system as well works shared between more encephalized primates (including (Bali and Kovacs, 2003; Blevins et al., 2003; Hoyda and Ferguson, humans) and lower mammals. 2010; Hrabovszky and Liposits, 2008). 34 brain research 1580 (2014) 22–56

recursive control systems within and outside the knot ((Brown deeper understanding of the pharmacodynamics, pharmacoki- et al., 2013; Kitano, 2004)and(Table 3)) are important in terms of netics and mechanism of action of IN OT. In terms of pharma- our global understanding of OT function, and highlight that cological development, though treatment of anxiety disorders is bowtie architectures do not depict simple left-to-right informa- typically chronic (multiple doses over weeks) very few clinical tion flow. Rather, extensive local and global feedback regulations trials of chronic OT in humans with anxiety disorders have been are imposed at each step in the system. Regulation, then, occurs performed (Table 2). The few studies that have followed this both within the central knot and between left and right sides (as pharmacological standard of care have generated initial positive explicated, for example, in Feldman's biobehavioral synchrony results in generalized anxiety disorder (Feifel et al., 2011)and model (Feldman, 2007; Feldman et al., 2013)). negative results in obsessive compulsive disorder (den Boer and Conceptually, a bowtie model of OT provides a framework Westenberg, 1992; Epperson et al., 1996). for future translational, preclinical, and clinical OT research, Much more numerous than clinical trials are pharmaco- may aid our understanding the role of the OT system in a the logical studies examining single-dose OT as a pharmacologi- development and treatment of a variety of other disorders cal probe. These studies have been performed in a variety of (including addictions (McGregor and Bowen, 2012)), and may different models of anxiety, with different subjects (normals inform complex, interactive models of human development, and patients), and show positive, neutral, and negative attachment, and anxiety (Feldman, 2012; Milrod et al., 2014; effects on different parameters of anxiety (Table 2). From Nolte et al., 2011). Conceptualizing the OT/social neuropep- our perspective—based on the data in this review—anxiety tide system as a shared knot between Figs. 2a and b under- disorders worthy of exploratory proof-of-concept randomized scores areas of overlap between the diverse literatures on clinical trials with chronic OT include posttraumatic stress human attachment, fear, anxiety disorders and OT (Belzung disorder, social phobia, generalized anxiety disorder, and and Philippot, 2007; Esbjorn et al., 2012; Milrod et al., 2014; panic disorder. The converging fields of human OT research Nolte et al., 2011; Porges, 1998; Warren et al., 1997). that support this contention include: (1) studies of genetic variants of the OT system; (2) studies of OTs impact on aspects of anxiety (i.e. hormone levels, neurobiological 4. Human research: OT and anxiety responses) in normal subjects; (3) studies using OT in labora- tory models of anxiety disorders; (4) studies of OT in anxiety- Above, we reviewed problems with the term “anxiety”,a disorders populations; (5) studies of OT in disorders in which phylogenic and evolutionary perspective on OT's role in anxiety symptoms play a role (eating disorders, borderline anxiety, translational OT research, and a conceptual model PD); (6) studies of the role of the OT system in intermediate of the OT system's function as a biological ‘bowtie’. Each of behavioral phenotypes associated with anxiety. Auspicious these related areas inform the next section: the direct study signals from each of these arenas are reviewed below. of OT's role in human fear and anxiety. Though research on the human central OT system has mushroomed, a host of technical limitations still constrain our deeper understanding of its development, dynamics, and func- tion. These include: (1) lack of a radioligand for OTR; (2) a limited understanding of functional role of OT genetic variants; (3) lack a (footnote continued) biological activity (i.e. activate the OT receptor), data on this have centrally-active OT antagonist for human research; (4) lack of a not been published. As such, radioimmunoassay (RIA) on clear understanding of the connection between central OT extracted plasma (Zhang et al., 2011) is currently the consensus 4 system activity and peripheral OT levels ;and(5)lackofa plasma assay of choice (Paul Zak, personal communication); (3) in humans, pOT and CSF OT levels—at least at baseline state—do 4In discussing the role of OT in anxiety, several significant not correlate (Kagerbauer et al., 2013) (but see Wang et al., 2013), caveats constrain conclusions supported by or derived from and CSF OT levels best reflect central activity and release studies of OT levels, whether those levels are measured in CSF, (Veening et al., 2010). This does not imply, though, that pOT plasma (pOT), urine or saliva (sOT). As discussed in a growing, levels are irrelevant to brain or behavior. Peripherally-delivered cautionary literature on this topic (Kagerbauer et al., 2013; OT—which technically does not cross the BBB—has behavioral McCullough et al., 2013; Szeto et al., 2011; Young and Anderson, effects in animals (Feifel and Reza, 1999; Feifel et al., 2012) and 2010), a host of assay validity issues in the measurement of OT humans (Hollander et al., 2003), perhaps mediated through the mean that we “know less about the neurobiological and beha- vagal nerve (Porges and Carter, 2012); (4) correlations between vioral significance of peripheral levels of OT” than many reports sOT levels and pOT levels do not approach those in other imply (McCullough et al., 2013). Concerns around the measure- endocrine literatures (i.e. salivary cortisol or testosterone) (see ment of OT levels include the fact that: (1) pOT levels differ McCullough et al. (2013) for references); (5) compared with urinary significantly based on sample processing: levels from extracted OT, pOT and sOT levels, relatively few papers examining CSF samples often differ from unextracted samples by a factor of 10 levels of OT in humans have been published (see Section 4.4) (see McCullough et al. (2013) for references), and levels from (Clark et al., 2013; Jokinen et al., 2012; Kagerbauer et al., 2013; Lee extracted and unextracted samples do not correlate with each et al., 2009; Sasayama et al., 2012; Wang et al., 2013). Summing up other (Szeto et al., 2011); (2) immunoassay methods like the this important aspect of the OT field, McCullough et al recently commonly-used enzyme immunoassay (EIA) are based on the opined that “until methods for measuring OT in any human fluid assumption that the target antibody is specific for only the are validated and standardized, interpretation of reported results analyte of interest (i.e., OT). In the case of the most-commonly remains murky at best” (McCullough et al., 2013). In the context of used OT EIA, this assumption is dubious given that OT- this review, these data encourage significant caution in inter- degradation products may cross-react with the antibody (Szeto preting the relationship between OT levels and anxiety states, et al., 2011). Though these OT degradation products may have traits and diagnoses. brain research 1580 (2014) 22–56 35

4.1. Genetic and epigenetic variations in OT systems and importance of this form of social signaling, for example, anxiety fearful faces consistently activate the amygdala (Adolphs, 2008), a hub of the brain's salience and danger-detection Like most psychiatric disorders, a substantial amount of the system, and known anxiolytics dampen these responses variance in the incidence of anxiety disorders is explained by (Blair and Curran, 1999; Del-Ben et al., 2012; Paulus et al., genetic factors (Norrholm and Ressler, 2009). Speaking to the 2005; Zangara et al., 2002). Within this context, some of IN novelty of this area of research, however, a recent review of the OT's effects on brain response to certain facial emotions topic did not mention OT (Norrholm and Ressler, 2009). This will (downregulating amygdala activity) seem to indicate an surely change, for in recent years, a large number of studies have anxiolytic effect, though mixed results are found, with other examined the relationship between common allelic variations in studies showing amygdala activation with social stimuli the OT system and a number of parameters of interest to anxiety (Domes et al., 2013a; Zink and Meyer-Lindenberg, 2012). disorders, including anxiety-related personality traits, brain Facial processing tasks reveals a fundamental conceptual structure, brain dynamics and responses to adversity and stress challenge in interpreting this aspect of the OT literature: a ((Bradley et al., 2011; Love et al., 2012; Rodrigues et al., 2009; diathesis between OT increasing social salience and its role in Thompson et al., 2011; Tost et al., 2010; Wang et al., 2013a,b)and decreasing social anxiety. As discussed in Section 1.3, social Brune (2012) and Kumsta and Heinrichs (2013) for review). Due to stimuli are inherently ambiguous, and can be a signal of both space constraints, we will not review these studies here (see safety and danger. The degree to which OT has a nonspecific Bakermans-Kranenburg and van Ijzendoorn (2013b) and Kumsta effect (increasing social salience in general) verses its effects and Heinrichs (2013) for recent reviews). on the processing of stimuli with a certain valence (i.e. happy Also important as regards the role of the OT in the develop- vs. fearful faces) is an active area of investigation (Domes ment of anxiety disorders are converging findings that the OT et al. (2014), Marsh et al. (2010), Prehn et al. (2013), Radke et al. system contains a degree of programmable, developmental- (2013), Shahrestani et al. (2013) for reviews, see Shahrestani window specificplasticityconferredviaepigeneticchanges et al. (2013) and Van IJzendoorn and Bakermans-Kranenburg (Champagne, 2008; Kumsta et al., 2013; Mamrut et al., 2013; (2012). Murgatroyd and Spengler, 2011; Stolzenberg et al., 2012; A last important distinction in this research arena is that Unternaehrer et al., 2012; Wang et al., 2013). These although certain experiments have suggested OT has a developmentally-canalized changes may occur at many levels benzodiazepine-like effect (de Oliveira et al., 2012a, 2012b; of the system (Table 3), including OT's role in regulation of the Pedersen et al., 2013), benzodiazepines cause impaired facial HPA axis and its role in amygdala-anchored brain systems (see recognition (Coupland et al., 2003), but see Kamboj and Section 4.7.3). We cannot do justice to this rich and rapidly Curran (2006) and Murphy et al. (2008). Serotonergic agents, developing literature, but recommend Kumsta et al. (2013), Nolte on the other hand, often improve facial recognition, though et al. (2011) and Veenema (2012) for recent reviews. these effects depend partly on duration of treatment ((Murphy et al., 2009; Norbury et al., 2009), and see 4.2. IN OT in healthy humans Macdonald et al. (2013b) for discussion of clinical implications of OT and facial processing). Avarietyofanxiety-relevantstudiesinbothhealthyand To summarize this important area, though it is clear that psychiatrically ill humans using IN OT have been performed IN OT has an effect on the visual processing of human faces (Table 2). These studies have examined OT's role in the proces- (Shahrestani et al., 2013), and though these findings may sing of disorder-specificstimuli(i.e.socialapprobationinsocial have relevance to OT's role as a therapeutic anxiolytic phobia) and more general threat-relevant stimuli (i.e. fearful and agent, specificdetailsoftheseassociationsareyettobe angry faces), as well as IN OT's action within a number of elucidated. anxiety-related laboratory models. Using clinical anxiety disor- ders as our point of reference, we note significant variation in the association between different experimental factors linked to OT 4.2.2. IN OT, anxiety, and social stress (i.e. amygdala hyperactivity, cortisol elevation, autonomic reac- Besides static facial expressions, active social approbation tivity), anxiety disorders, and their treatments (Craske et al., and evaluation are potent activators of the stress response 2009; Vreeburg et al., 2010; Watson, 2005). (Dickerson and Kemeny, 2004), and several studies evaluate OT's impact in these situations. Using a laboratory model of 4.2.1. OT, facial stimuli, and anxiety public speaking called the Simulated Public Speaking Test Many short-term studies of IN OT have examined its impact (SPST) (McNair et al., 1982), de Oliveira et all demonstrated on the processing of facial stimuli in normal populations that IN OT (1) significantly decreased subjective anticipatory (Shahrestani et al. (2013), Van IJzendoorn and Bakermans- anxiety and increased subjective sedation before the speak- Kranenburg (2012) for recent reviews). Face processing stu- ing task; as well as (2) decreased skin conductance at most dies relate to human anxiety because human facial expres- time points throughout the experiment (de Oliveira et al., sions—especially eye-region related signals—are evolved, 2012a, 2012b). Notably, the effect of OT was similar to the reflexively-processed displays of the emotional and motiva- effect of diazepam in the SPST model (Zuardi et al., 1993). In tional states of others. As such, they reliably capture atten- an earlier seminal experiment in this literature, Heinrich et al tion and activate fear-related brain systems and behaviors in demonstrated that both OT and social support exerted sub- observing others (Adolphs 2008, 2009; Todorov, 2008; jective anxiolytic effects and attenuated cortisol responses to Tomasello et al., 2007). Consistent with the evolutionary a social stressor (Heinrichs et al., 2003). 36 brain research 1580 (2014) 22–56

4.2.3. OT in conditioned fear experiments 4.2.5. Oxytocin, anxiety and the HPA axis Fear learning and fear extinction play in important role in the Due to the role of the HPA system in stress responses development and treatment of several anxiety disorders (Carrasco and Van de Kar, 2003; Chrousos, 1995; Selye, (Bishop, 2007; Indovina et al., 2011; Mineka and Zinbarg, 1936), activation of this system is considered an intrinsic part 2006). Specifically in the case of PTSD, prolonged exposure of mammalian defense systems, anxiety and fear. HPA therapy (with the expectation of extinction learning of con- activation is partially regulated by tonic inhibitory control ditioned fear responses) is one of the preferred treatments and negative feedback from both the hippocampus and (Foa, 2006). In translational OT research, conditioned associa- prefrontal cortex (Brake et al., 2000; Chrousos, 1995; Diorio tion experiments form part of the backbone of studies et al., 1993; Herman and Cullinan, 1997). OT, notably, has examining the acquisition, learning and extinction of anxiety interaction with several critical components of this axis (Milad et al., 2006; Myers and Davis, 2007). This research has including the BNST and PVN (the source of corticotropin- demonstrated that OT plays a significant role in conditioned releasing hormone (Chrousos, 1995)) and the hippocampus association and fear learning (Lukas et al., 2013), partly based (Stoop, 2012)(Fig. 3). Translational animal research on OT's on its role in the amygdala (Knobloch et al., 2012; Petrovic effects on the HPA axis have clearly demonstrated its mod- et al., 2008; Viviani et al., 2011; Viviani and Stoop, 2008). ulatory role (Amico et al., 2008; Neumann, 2002; Slattery and Several groups have examined the role of IN OT in human Neumann, 2010), and these endocrine effects are often cited fear learning. In an important experiment, Petrovic et al as evidence of OTs inherent anxiolytic activity (Engelmann found OT given shortly after acquisition of conditioned et al., 2004; Slattery and Neumann, 2008). exposure abolished conditioned fear responding via down- Outside of animal models (which allow direct access to the regulation of amygdala activity (Petrovic et al., 2008). Two central arm of the HPA axis), a large number of studies in a other trials using a fear-potentiated startle experimental variety of normal and psychiatric populations have examined IN model found that OT exerted a facilitative effect on extinction OT's effects on cortisol levels, with mixed results that will not be recall (but not extinction) (Acheson et al., 2013), and that reviewed in detail here (Burri et al., 2008; Cardoso et al., 2013; de although OT did not increase startle to cued threat, it did Oliveira et al., 2012a, 2012b; Ditzen et al., 2009; Heinrichs et al., have the “anxiogenic” effect of increasing potentiated startle 2003; Linnen et al., 2012; Quirin et al., 2011; Simeon et al., 2011). responses to unpredictable threat (Grillon et al., 2013). These Several caveats about the effect of IN OT on cortisol in humans, somewhat mixed results underscore two important points however, are warranted. First, peripheral markers of HPA axis about the therapeutic role of OT. First, a simple “amygdala activity (i.e. salivary or plasma cortisol levels) have both practical dampening” model of OT's activity needs revision (see Bartels and theoretical limitations, including timing factors, stressor- (2012) for discussion); and (2) careful examination of the related factors, and different HPA responses in different patient timing of OT exposure relative to cue presentation as well populations (Dickerson and Kemeny, 2004; Hellhammer et al., as the context of the experiment and nature of the cues 2009). In terms of relevance to clinical treatment, though there is (social vs. nonsocial) in these trials is critical (Acheson et al., apredictablelinkbetweencertaintypesofstressorsandHPA 2013; Lahoud and Maroun, 2013; Toth et al., 2012). Recent activation in normal humans (Dickerson and Kemeny, 2004; concerns regarding the shortcomings of the fear extinction Koolhaas et al., 2011), the link between human anxiety disorders model of human anxiety disorders have recently been raised and cortisol levels is variable (Hek et al., 2013; Vreeburg et al., (Milrod et al., 2014), and are also germane here. Finally, it is 2013, 2010). Complexity is also introduced by evidence that worth noting different correlations between fear-related con- stress-induced cortisol levels may influence the release of OT ditioning experiments and individual anxiety disorders, with (Pierrehumbert et al., 2012; Tops et al., 2012), and data indicating most consistent results in PTSD (Craske Rauch et al., 2009). complex relationship between acute stressors and changes in This may indicate that the role of OT in PTSD may differ from endogenous cortisol and OT levels (Tabak et al., 2011; Taylor that in other anxiety disorders (Olff et al., 2010). et al., 2010). A final important point regarding of OT's interaction with the 4.2.4. IN OT in other models of anxiety HPA axis is that dysregulation of this axis and its consequences

Inhalation of 7.5 % CO2 is an anxiogenic stimulus which —impaired plasticity and neurogenesis—are one of the plausible serves as a preclinical model of generalized anxiety (Bailey mechanisms mediating the link between traumatic early experi- et al., 2007). CO2 inhalation increases heart rate, blood ence and later psychopathology (Apter-Levy et al., 2013; Heim pressure and subjective feelings of anxiety, fear and tension et al., 2010; Kendler et al., 2004; McCrory et al., 2012). For (Bailey et al., 2005), and these responses are attenuated by example, altered amygdala-prefrontal connectivity found in both lorazepam and paroxetine (Bailey et al., 2007). In an humans exposed to maternal deprivation is mediated by cortisol experiment using this model, de Oliveira et al demonstrated levels (Gee et al., 2013). These data and others mentioned that both OT and lorazepam attenuated the anxiogenic throughout this review suggest that HPA-OT interactions may fi response to CO2 (de Oliveira et al., 2012a, 2012b). Speci cally, be a mediator of some of the effects of early adversity on the although physiological and hormonal anxiolytic effects were connectivity and development of key brain structures related to not seen in this experiment, OT significantly impacted sub- anxiety (Elton et al., 2013). This point also relates to psychiatric jective anxiety symptoms. therapeutic agents, given that modulation of the HPA axis in To summarize: OT has demonstrated anxiolytic-type considered an important component of the therapeutic action of effects in a number of preclinical human models, though the SSRIs (see Hesketh et al. (2005) for discussion). some results in this area have been lukewarm and even To summarize: though significant preclinical literature seemingly contradictory (Grillon et al., 2013). indicates that OT-HPA interactions are important and that brain research 1580 (2014) 22–56 37

Fig. 3 – Brain nodes and networks implicated in OT's effect on anxiety symptoms and disorders in humans. Key nodes, based on both neuroimaging data using IN OT and OTR staining in humans include: (1) amygdala and bed nucleus of the stria terminalis (BNST) (left and right pullout); (2) hypothalamic-pituitary (HPA) axis (right pullout); (3) prefrontal-amygdala connectivity networks associated with regulation, second-order representation, and integration of information (left pullout and asterisks). Left pullout highlights key nodes and connections of an amygdala-anchored network as well as the distinction between fear—a phasic alarm response to present or imminent pain or danger—and anxiety—a more chronic, future-oriented anticipatory state (see text). Cortical areas with reciprocal connections to amygdala are on bracketed on left side of pullout, brainstem areas on the right. Underlined structures are those recently identified as containing OTR in cells or fibers in humans (see Boccia et al. (2013) for full list). Right pullout emphasizes BNST-PVN connections as key regions for crosstalk between OT and the HPA axis (reprinted with permission from Dabrowska et al. (2011)). Asterisks indicate areas identified in both neuroimaging experiments using IN OT and in anxiety disorders. Abbreviations:; ACC-anterior cingulate cortex; AMY- amygdala; AVP-arginine vasopressin; BLA-basolateral nucleus of the amygdala; BNST-bed nucleus of the stria terminalis; CeL/CeM-lateral/medial portion of the central nucleus of the amygdala; CRF-corticotrophin-releasing factor; DVC-dorsal vagal complex; GABA-gama-aminobutyric acid; GLUT-glutamate; HC-hippocampus; NRPc-nucleus reticularis pontis caudalis; OTR- OT receptor; PAG-periaqueductal grey; PFC-prefrontal cortex; PVN-paraventricular nucleus; SON-supraoptic nucleus. For references, see text and (Bethlehem et al., 2013; Boccia et al., 2013; Davis et al., 2010; Huber et al., 2005; Stoop, 2012; Viviani et al., 2011; Viviani et al., 2010; Zink and Meyer-Lindenberg, 2012).

IN OT may regulate this axis in certain states and human one of the more important roles of psychopharmacology in subjects, the relevance of these findings to the development this condition (Olff et al., 2010). and treatment of human anxiety disorders is still uncertain. Given its role in social learning and certain forms of prosocial behavior, social phobia (SOP) or social anxiety disorder is an auspicious target for OT. Though no chronic studies in SOP have 4.3. IN OT in psychiatric populations with anxiety been done, mixed results from single-dose trials exist. In one study, single-doses of OT were not found not to augment Very few treatment studies have been done in patients with cognitive therapy for this disorder (save an improvement in anxiety disorders (Table 2). In one of the first, Pitman et al self-rated performance) (Guastella et al., 2009), whereas imaging found that OT reduced physiological responding (HR, GSR, studies in patients SOP found OT reduced hyperactive amygdala facial EMG) to personal combat imagery prompts (Pitman responses (Labuschagne et al., 2010)andreducedhyperactive et al., 1993). More recently, functional imaging studies in medial frontal and anterior cingulate responses to emotional PTSD patients indicate that OT reduces amygdala activation faces (Labuschagne et al., 2011). during the presentation of fearful and angry faces (Koch et al., In one of the only double-blind, placebo-controlled studies of 2013). Studies of fear learning in normals (reviewed above) chronic OT in generalized anxiety disorder, Feifel et al gave 13 are also important in terms of the use of OT in PTSD, as subjects twice-daily IN OT for 3 weeks, and demonstrated a pharmacological augmentation of exposure therapies may be trend-level dose-by-gender effect such that males treated with 38 brain research 1580 (2014) 22–56

OT showed a significant improvement in HAM-A scores at week examining the effects of OT on anxiety and anxiety-related 2, whereas females showed a tendency for increased anxiety symptoms in other populations are relevant to the larger (Feifel et al., 2011). The indication of a sex-dependent effect in question of the role of OT in anxiety. One population that this study is critical to follow up, given significant, converging has been studied in this context is borderline personality data that sex is a critical individual variable in understanding the disorder (BPD). Though BPD patient's affective experience human OT system (Macdonald, 2012). Interestingly, recent work includes irritability, dysphoria, and rage, many patients do in invertebrates indicates that sex-specificOTeffectsextendas suffer with disorganizing levels of anxiety and fearfulness far back as the function of an OT-like precursor (nematocin) in (Leichsenring et al., 2011). Overall, mixed findings with OT in Caenorhabditis elegans,wheresex-specificeffectsonmating BPD exist, with single dose-studies demonstrating that IN behavior in males are found (Garrison et al., 2012). OT impedes trust and prosocial behavior (Bartz et al., 2011), Ironically, at the time of this review, the largest anxiety- attenuates post-stressor dysphoria (Simeon et al., 2011), and disordered patient group treated with chronic OT is patients normalizes amygdala hyperactivity in responses to angry with OCD (den Boer and Westenberg, 1992; Epperson et al., faces (Bertsch et al., 2013). Importantly, effects in several of 1996), a disorder no longer categorized in the DSM anxiety these studies were moderated by attachment parameters disorder grouping. The two negative studies in this condition and early traumatic experience. These studies, though demarcated the start of significant pause in the clinical short-term, have generated interest in OT as a chronic investigation of OT as a human anxiolytic. treatment for BPD. Anxiety is also an important feature of eating disorders, and 4.4. OT levels in patients with anxiety disorders the OT system is intercalated withcentralsystemsregulating ingestion and weight (Valassi et al., 2008; Wu et al., 2012). Several Any discussion regarding OT levels must consider the sig- studies have related pOT levels to different aspects of anorexia nificant caveats discussed in footnote 3. With these concerns nervosa (AN) (Lawson et al., 2013, 2011). Lawson et al recently in mind, several studies have indicated that endogenous OT reported that elevated subjective anxiety (STAI) levels were plasma levels (pOT) are related to prosocial behaviors associated with elevated mean, nadir, and peak pOT in AN (Grewen et al., 2005; Light et al., 2005), and that lower OT patients, and suggested that up to 51 % of anxiety symptoms levels occur in disorders with prominent social dysfunction (i. (STAI trait) were directly attributable to OT secretion (Lawson e. autism (Modahl et al., 1998), schizophrenia (Keri et al., 2009; et al., 2013). Besides BPD and AN, patients with the rare, Rubin et al., 2010), but see also Rubin et al. (2013)). Addition- inherited disorder called fragile X syndrome often manifest ally, a mixed literature has shown that pOT levels vary as the anxiety, and a small pilot study showed that IN OT improved result of stress, depending on individual and context factors certain parameters of social anxiety (eye gaze) and lowered (Light et al., 2005; Taylor et al., 2006; Turner et al., 1999). Few cortisol levels (Hall et al., 2012). studies of OT levels in anxiety disorders exist. In OCD, In summary: OT and aspects of the OT system have been patients were found to have higher CSF OT levels in one associated with anxiety and anxiety-related states in several study (Leckman et al., 1994), but not a follow up (Altemus human psychiatric disorders, but some of these findings are et al., 1999). Though an earlier study found no significant mixed and moderated by individual factors as well as con- differences between pOT in patients with generalized social cerns about the relevance of pOT levels (footnote 3). anxiety disorder and normals (Hoge et al., 2008), within the social anxiety group, pOT levels were associated with symp- tom severity. In a later study, Hoge et al reported decreased 4.6. The role of OT in anxiety-related behavioral pOT levels in patients with generalized social phobia during a phenotypes laboratory task of prosocial behavior (Hoge et al., 2012). Given the role of childhood trauma in the development of anxiety Apart from categorical anxiety disorders, a final important fi disorders (Milrod et al., 2014; Nolte et al., 2011), ndings from arena that informs our understanding the role of OT in a small study of 22 abuse victims showing inverse association human fear and anxiety is OTs connection with anxious between CSF OT concentrations and abuse are of interest temperaments and personality traits. These more stable, (Heim et al., 2009). Altogether, extant literature examining OT chronic characteristics are measured using a variety of levels in both normal and certain clinical populations (i.e. instruments including behavioral observations and self- anorexia (Lawson et al., 2013), autism (Miller et al., 2013)) reports5 . They include (1) anxious temperament (aka “inhib- supports the idea that the peripheral release of OT system ited temperament”); (2) attachment anxiety; (3) trait anxiety; may be part of an acute stress response or anxiety states. and (4) harm avoidance (see Table 1). Sometimes considered Although there have been connections made between tem- peramental aspects of anxiety and pOT (Weisman et al., 5Importantly, though self-reports assay an important para- 2013a), direct correlations between anxiety disorders and OT meter of anxiety disorders, dissociation between physiological levels is minimal. symptoms and self-reported anxiety (i.e. response dyssynchrony) is frequently reported in anxiety studies (Grillon et al., 2008; 4.5. IN OT and OT levels in other populations with anxiety Hodgson and Rachman, 1974; Kemp et al., 2004; Ruys and Stapel, 2008; van Honk et al., 2005). A common distinction in this area assigns nonconscious manifestations of the anxiety/fear system As mentioned above, anxiety is a transdiagnostic symptom to subcortical structures (i.e. amygdala, hypothalamus) and con- and forms a clinical focus of treatment in several disorders sciously perceived manifestations to cortical activity (Morris outside of the codified anxiety disorders. As such, studies et al., 1999; Ribary, 2005; Ruys and Stapel, 2008). brain research 1580 (2014) 22–56 39

endophenotypes6 or intermediate phenotypes for anxiety much more likely to be highly anxious) (Weisman et al., disorders (Schwartz et al., 2012), anxiety-related personality 2013a), whereas pOT levels were inversely correlated with traits are predisposing factors for both depression and trait anxiety scores in breastfeeding women (Stuebe et al., anxiety-related disorders (Beesdo et al., 2010; Farmer et al., 2013). Lastly, in a neuroimaging study illuminating the 2003; Ono et al., 2002; Rosenbaum et al., 1993; Starcevic and dynamic relationship between OT and dopamine systems, Uhlenhuth, 1996), supporting the “internalizing disorder” and female C allele carriers of the rs4813625 allele of the OT RDoC negative valence categorization schemes discussed in receptor (OTR) had higher stress-induced dopamine release, Section 2. As they are related to the development of anxiety higher trait anxiety scores and higher attachment anxiety disorders and have links with OT, these data are worth (Love et al., 2012). These findings echo reports finding ele- considering in the broad spectrum of OT-anxiety research. vated pOT levels in distressed women (Taylor et al., 2006, Afrequentlyexaminedanxiety-relatedtemperamentalvar- 2010), and reports correlating pOT, trait anxiety and early life iation is inhibited temperament or “anxious temperament”. stress (Opacka-Juffry and Mohiyeddini, 2012). Though this particular phenotype has not yet been directly Harm avoidance (HA) is a temperamental trait similar to related to OT in humans, experiments in rats demonstrate that trait anxiety, but measured using a different self-report, only females with anxious temperament demonstrate altera- Cloninger's TCI (Cloninger, 1987). Like trait anxiety, HA is tions in AVPR-1a levels in the amygdala when subjected to associated with being cautious, apprehensive and pessimistic male aggression (Poirier et al., 2013). Moreover, anxious tem- and with anxiety disorders (Ball et al., 2002; Lochner et al., perament negatively correlates with prosocial traits like extra- 2007; Starcevic and Uhlenhuth, 1996; Wachleski et al., 2008). version and cooperativeness (Zukerman and Cloninger, 1996), HA is also associated with a variety of laboratory measures which have been related to OT levels and brain responses to IN (i.e. startle, endocrine responses, functional and structural OT (Andari et al., 2012; Groppe et al., 2013). This behaviorally- brain responses) connected with anxiety disorders (Corr et al., anchored temperamental variation—vital in our understanding 1997; Tyrka et al., 2008; Van Schuerbeek et al., 2011; Yamasue of the development of human anxiety disorders (Biederman et al., 2008; Baeken et al., 2009; Li et al., 2012; Westlye et al., et al., 2001; Hirshfeld et al., 1992; Schwartz et al., 2012)—is an 2011; Yang et al., 2009). Wang et al. studied the rs53576 interesting target for future OT research. variant of the OTR gene and found that HA was associated Whereas inhibited temperament is measured using beha- with this gene in females (homozygous A-allele had higher vioral observations, the personality characteristic called trait HA than G-allele carriers), and also that this gene variant was anxiety is measured with the state-trait anxiety inventory associated with significantly reduced gray matter volume in (STAI) (Spielberger, 1983). Trait anxiety describes a stable bilateral amygdala and reduced functional connectivity tendency toward tension, anxiety, and negative affectivity, between amygdala and dlPFC (Wang et al., 2013a,b). In an and is associated with anxiety disorders and functional earlier study, the same OTR gene was found to exert an effect abnormalities in amygdala and prefrontal cortex (Barrett on interregional coupling of amygdala during emotionally and Armony, 2009; Bishop, 2009; Indovina et al., 2011; salient social cues (Tost et al., 2010). Moreover, G x E studies Otero-Garcia et al., 2013; Sehlmeyer et al., 2011). Phenomen- suggest that this gene or other OTR genes may function as ologically, high trait anxiety and negative affectivity are factors in resilience to psychopathology (Bradley et al., 2013; common to all anxiety disorders (Clark et al., 1994; Watson, Brune, 2012). In a second study of the common OXTR variants 2005). In terms of relationships with OT, trait anxiety was rs237900 and rs237902, a genotype-by-sex interaction was found related to pOT levels in a sex-dependent manner in found to explain 11 % of the variance between levels of HA in one study of men and women (trait anxiety was lower among females (Stankova et al., 2012). Overall, these findings suggest men with higher OT, and women with very high levels were a possible mechanistic pathway from OTR genotype to anxiety-related traits (HA) to anxiety disorders, through modulation of prefrontal-amygdala coupling. A final trait measure of interest in relation to OT and 6As part of the goal of bridging the “gene-to-phene” gap, anxiety is attachment anxiety, a measure of sensitivity to or neuroscientists have searched for stable, reliably measurable traits called endophenotypes (from the Greek ‘endos’ or ‘with- worry about rejection or loss of relationships with intimates in’ ‘phainein’, ‘to show’)(Bearden and Freimer, 2006; Gottesman measured using the experiences in close relationships ques- þ and Gould, 2003; Waldman et al., 2006). In its most liberal tionnaire (ECR-R) (Brennan et al., 1998). In several studies, definition, key characteristics of an endophenotype include a attachment anxiety (and the related parameter of attachment fi quality that is (1) quanti able; (2) associated with genetic risk for avoidance) have been found to moderate responses to IN OT a disorder; (3) abnormal in both patients and probands; and (4) (Bartz, 2012; De Dreu, 2012) and predicted trust-related OT relatively invariant (Bearden and Freimer, 2006). This definition includes certain behaviors, self-report measures and cognitive responses (Kiss et al., 2011). Furthermore, attachment anxiety processes present in nonclinical populations, and therefore in women has been correlated with higher pOT levels occurring “under the observable illness phenotype”. Others have (Weisman et al., 2013a). In a study mentioned above, C allele called these characteristics “intermediate phenotypes” (Stein carriers of the rs4813625 C allele exhibited higher attachment et al., 2008), “exophenotypes” (Meyer-Lindenberg and Weinber- anxiety (Love et al., 2012). A caveat to the relationship of “ ger, 2006) or in the case of anxiety disorders, anxiety spectrum attachment anxiety and anxiety disorders is that in a large phenotypes” (Hettema et al., 2008). A more restrictive definition sample of psychiatric outpatients, attachment anxiety corre- for endophenotype—more applicable to translational neu- roscience—opines that they should be laboratory-based biomar- lated more strongly with mood and personality disorder kers: processes occurring underneath observable/reportable symptoms than harm avoidance or anxious temperament behavior (i.e. Light et al., 2012). (Macdonald, 2012), so the clinical significance of these 40 brain research 1580 (2014) 22–56

findings in regards to the effects of IN OT in psychiatric perspective, a variety of studies indicate that anxiety dis- disorders is unknown. orders are associated with dysfunctional connectivity of In summary: a review of the growing literature linking amygdala with prefrontal and brainstem regions, as well as aspects of OT to anxiety-related temperamental variations aberrant amygdala connectivity within the central executive supports the hypothesis that genetic variations in the OT and salience networks (Blair et al., 2012; Campbell-Sills et al., system may play a role in the neurobiological substrates of 2011; Damsa et al., 2009; Etkin et al., 2010, 2009; Etkin and anxiety-related predispositions, and—with caveats about OT Wager, 2007; Kim et al., 2011; Kim and Whalen, 2009; Martin levels (footnote 3)—that these levels may correlate with et al., 2009; Menon, 2011; Shin and Liberzon, 2010; Sylvester certain anxiety-related traits. et al., 2012).

4.7. Neural substrates for OT and anxiety: nodes and 4.7.2. Periaqueductal gray networks The periaqueductal gray (PAG) is a brainstem structure involved in fight-or-flight reactions to a proximal threat In this final section, we review both putative and proven (Panksepp, 1998), sexual and maternal behaviors (Salzberg neural substrates that may mediate some of OT's anxiety- et al., 2002), anxious temperament (primates) (Fox et al., related effects (Fig. 3). Though space does not permit a review 2008), dread, threat imminence, sympathetic response to of the similarities and differences in the brain regions socioevaluative threat, and negative emotion (in humans) implicated in different anxiety disorders (see Menon (2011) (Buhle et al., 2013; Craske et al., 2009; Hermans et al., 2012; and Sylvester et al. (2012) for recent reviews), we highlight Linnman et al., 2012; Mobbs et al., 2007; Tasker, 1982; Wager several key nodes and networks that inform the putative role et al., 2009). The role of the PAG in anxiety and defensive of OT in anxiety disorders and warrant further investigation. responses has been much more well-studied in animals than humans, due to technical limitations in human neuroima- 4.7.1. Networks and hubs ging, which have emphasized amygdala and cortical struc- Our understanding of fear and anxiety disorders has been tures (but see Buhle et al. (2013), Damasio et al. (2000), Mobbs significantly influenced by systems neuroscience and the et al. (2007, 2010), Wager et al. (2009) and Linnman et al. (2012) segregation of the human brain into functional networks: for review). In an interesting recent study of relevance to OT, large and small-scale brain regions that demonstrate intrinsic eye contact in people with PTSD from interpersonal trauma functional connectivity (Menon, 2011; Seeley et al., 2007; stimulated increased activation within the superior colliculus Sylvester et al., 2012). These networks have critical “nodes” (SC) and periaqueductal gray (PAG) (Steuwe et al., 2014), both or “hubs”, defined by both intrinsic properties as well as regions associated with OT activity (Gamer et al., 2010; Ge extrinsic connectivity (Honey et al., 2009; Passingham et al., et al., 2002; Yang et al., 2011). Regarding OT's potential role in 2002). Neural hubs coordinate and mediate alterations in and human anxiety, OT is found in the PAG of humans (Jenkins between different networks over multiple time scales (Honey et al., 1984), though OTR are not (Boccia et al., 2013). Animal et al., 2007; Sporns et al., 2007). Regions and nodes critical to studies indicate that PAG has a role in some aspects of OT's an understanding of OT's role in fear and anxiety include (1) regulation of the sympathetic response and anxiety-related subcortical and brainstem structures which are part of the behavior (Figueira et al., 2008; Viviani et al., 2011), as well as mammalian defense system (i.e., periaqueductal gray (PAG), pain modulation (Ge et al., 2002; Tracey et al., 2002; Yang hypothalamus, amygdala); (2) prefrontal cortical regions et al., 2011). Though the role of the PAG in human anxiety associated with modulation, integration and regulation of states and disorders has not been well specified (but see these regions (i.e., anterior cingulate cortex (ACC) and other Buhle et al. (2013) and Hermans et al. (2012)), OT's effects in prefrontal cortical regions (vmPFC)); and (3) cortical regions paradigms where PAG has been shown to be activated associated with integration of internal body states or “inter- (Hermans et al., 2012; Mobbs et al., 2007; Steuwe et al., 2014) oceptive prediction” (i.e., anterior insula) (Damsa et al., 2009; would be informative. Deakin and Graeff, 1991; Etkin and Wager, 2007; Martin et al., 2009; McNaughton and Corr, 2004; Paulus and Stein, 2006; 4.7.3. Amygdala Shin and Liberzon, 2010)(Fig. 3). Elegant animal experiments The amygdala is a key node in mammalian fear networks using targeted delivery of OT agonists, antagonists, as well as (LeDoux, 2003), in functional studies of human anxiety dis- optogenetics have allowed functional dissection of the micro- orders (Etkin and Wager, 2007; Menon, 2011), and has fea- circuitry of OT's anxiolytic effects, especially in the amygdala tured prominently in both animal and human studies of the ((Huber et al., 2005; Knobloch et al., 2012; Viviani et al., 2011; short-term effects of IN OT (Bethlehem et al. (2013), Stoop Viviani and Stoop, 2008; Viviani et al., 2010) and Stoop (2012) (2012) and Zink and Meyer-Lindenberg (2012) for reviews). It is for review). In addition, a growing human literature docu- widely thought that along with insula, amygdala registers ments OT-mediated changes in functional activation of many and initiates a “first response” to a variety of fear-inducing of these brain regions in healthy controls (Bethlehem et al. stimuli as key nodes in what is called the salience network (2013) and Zink and Meyer-Lindenberg (2012) for reviews) and (SN) (which also includes dACC, substantia nigra, ventral in patients with anxiety disorders (Labuschagne et al., 2010, tegmental area) (Martin et al., 2009; Etkin and Wager, 2007; 2011). Menon, 2011; Paulus and Stein, 2006; Seeley et al., 2007). Therefore, though for organizational purposes we focus on Interestingly, despite its prominence in neuroimaging studies specific brain regions, we acknowledge the importance of a of humans treated with IN OT, until recently there was no functional-network level of analysis. From this latter direct evidence of OTR in human amygdala (but see Boccia brain research 1580 (2014) 22–56 41

et al. (2013)). Effects of IN OT on the human amygdala likely early trauma and deprivations have been associated with loss occur due to a combination of direct (wired) transmission (via of amygdala-based discrimination between mothers and direct projections from the PVN) as well as diffusion effects strangers (Olsavsky et al., 2013), amygdala size (Mehta et al., from somatodendritic OT release (Ludwig and Leng, 2006; 2009; Tottenham et al., 2010) and amygdala responses to Stoop, 2012). emotional faces (Dannlowski et al., 2013, 2012). In a seminal OT's role in human anxiety disorders is informed by study on the development of anxiety disorders, Burghy et al studies indicating that IN OT impacts amygdala-anchored found early life stress to be associated with increased cortisol networks in normals (Kirsch et al., 2005; Sripada et al., 2012) levels, resting-state amygdala-vmPFC connectivity and anxi- and normalizes hyperactive amygdala responses to visual ety symptoms 14 years later (Burghy et al., 2012). En toto, this social stimuli in social anxiety patients (Labuschagne et al., body of work informs human clinical research with OT, given 2010), patients with borderline personality disorder (Bertsch that early experience is an individual factor that appears to et al., 2013), and PTSD (Koch et al., 2013). These acute effects influence responses to IN OT (Macdonald, 2012). It has even of IN OT on amygdala parallel the effects of some known been hypothesized that the effect size of IN OT in certain anxiolytic agents like lorazepam (Paulus et al., 2005) and conditions may be mitigated by neglect of this developmental pregabalin (Aupperle et al., 2011), though other agents anxio- factor (Bakermans-Kranenburg and van Ijzendoorn, 2013a). lytic over the long term have short-term activating effects on amygdala (i.e. SRIs (Del-Ben et al., 2005; McKie et al., 2005)). 4.7.4. Bed nucleus of the stria terminalis Aside from the evidence from functional neuroimaging Besides amygdala proper, an underexplored area in terms of studies with acute IN OT, other links between OT, amygdala human anxiety disorders and OT is the bed nucleus of the and anxiety disorders comes from evidence that anxiety- stria terminalis (BNST), considered a functional part of the prone individuals and patients with anxiety disorders “extended amygdala” (Davis et al., 2010; Veinante and demonstrate decreased amygdala volume (Spampinato Freund-Mercier, 1997). Together, amygdala and BNST serve et al., 2009) and that genetic variations in OTR are related to as the primary output structures mediating the expression of amygdala volume in some studies (Inoue et al., 2011; Wang fear responses via connections to the brainstem and et al., 2013a,b, but see Tost et al. (2011)). Moreover, as hypothalamus (LeDoux et al., 1988; Oler et al., 2012; Stoop, discussed above, in females, variants of the common 2012)(Fig. 3). Experiments in both humans and animals rs53576 OTR gene affect (1) structural integrity in amygdala- disambiguating amygdala from BNST have supported the prefrontal connectivity pathways; (2) trait anxiety (harm distinction between amygdala—which mediates phasic fear avoidance); (3) resting-state connectivity between PFC and responses—and BNST, associated with the transition from amygdala (subsequently correlated with trait anxiety) (Wang brief responses to tonic, long-duration anxiety states (Davis et al., 2013a,b). Also worth noting are studies indicating that et al., 2010; Hammack et al., 2004; Lang et al., 2000; Pego et al., the structural and functional connectivity of the amygdala 2008; Ventura-Silva et al., 2012; Walker and Davis, 2008). and mPFC are central to anxiety disorders (Kim et al., 2011; Specifically, sustained input from the BLA to the BNST, Kim and Whalen, 2009), and that this functional connectivity mediated in part by corticotropin releasing hormone (CRF) is impacted by both OT (Sripada et al., 2012) and AVP (Zink (Lee and Davis, 1997), increases output from the BNST to the et al., 2010). hypothalamus and brainstem regions associated with sus- Crucially, in terms of the development of OT-targeted tained anxiety (Walker and Davis 2008)(Fig. 3). Also impor- therapeutics, we are not aware of any functional imaging tant in this context is that a population of hypothalamic PVN studies of the effects of chronically-dosed OT on amygdala OT neurons projects directly to the BNST (Buijs et al., 1980; function. Such studies are vital, given that certain premorbid Sawchenko and Swanson, 1983; Sofroniew, 1983), and human functional states have shown promise as biomarkers of post-mortem studies found OTs in the human BNST (Jenkins eventual therapeutic anxiolytic effects of a variety of anxio- et al., 1984) (but see Boccia et al. (2013)). Especially illuminat- lytic treatments (Bryant et al., 2008; Nitschke et al., 2009; Phan ing in this regard is Dabrowsa et al's recent examination of et al., 2013; Whalen et al., 2008a). the role of PVN OT neurons in reciprocal regulation of CRF- Taking a developmental perspective, alterations in OT induced anxiety states (and the HPA axis) via CRF neurons in systems and their modulation of amygdala may also inform BNST (see Fig. 1, R. pullout) (Dabrowska et al., 2011). This our understanding of the mechanisms wherein early trau- group has also noted that different populations of CRF matic experiences influences the development of anxiety and neurons exist in the PVN and BNST, the former glutamater- stress-related disorders (Apter-Levy et al., 2013; McCrory gic, the latter GABAergic (Dabrowska et al., 2013). et al., 2012; Richter-Levin and Maroun, 2010). The burgeoning, Despite the important role of the BNST, recent reviews of cross-disciplinary field studying these mechanisms high- human imaging studies using IN OT do not mention this lights the potential of OT research to inform questions about region (Bethlehem et al., 2013; Zink and Meyer-Lindenberg, the key mediators between early experiences and the devel- 2012). Some of the reason for this neglect is technical; the opmental of psychiatric disorders. Translational research has BNST (like the PAG) is small and difficult to identify (but see already demonstrated the role of OT in the intergenerational Alvarez et al. (2011), Choi et al. (2012), Mobbs et al. (2010), transmission of competent caregiving (Champagne et al., Somerville et al. (2010), Straube et al. (2007) and Yassa et al. 2001; Champagne and Meaney, 2001; Meaney, 2001; (2012)). In addition to technical challenges, much of the Pedersen and Boccia, 2002) and the role of the amygdala in literature examining brain responses to OT have used the development of attachment to traumatic caregivers short-term, evocative protocols, which by their nature are (Moriceau and Sullivan, 2005; Raineki et al., 2010). In humans, more likely to elicit amygdala than BNST. Recent imaging 42 brain research 1580 (2014) 22–56

studies utilizing advanced spatial resolution techniques have recent study by Phan et al. indicated that chronic SSRI recently validated the role of the BNST in sustained, antici- treatment of generalized SOP changed threat-related activity pation and monitoring of unpredictable threat vs. phasic fear in amygdala and vmPFC (Phan et al., 2013). Additionally, GAD in both trait-anxious and anxiety-disordered patients subjects show abnormal function of anterior cingulate (ACC) (Alvarez et al., 2011; Choi et al., 2012; Mobbs et al., 2010; (Blair et al., 2012; McClure et al., 2007), and in two studies of Somerville et al., 2010; Straube et al., 2007; Yassa et al., 2012). subjects with GAD, the degree of baseline activation of ACC For example, medication-free GAD patients in a high uncer- was positively correlated with the likelihood of response to tainty gambling task showed decreased amygdala activity anti-anxiety medication (Nitschke et al., 2009; Whalen et al., and increased activity in BNST (Yassa et al., 2012), and 2008b). individuals with greater trait anxiety show more activity in In regards to OT and prefrontal cortex, OTR mRNA is found BNST (and insula) when tracking threat proximity (Somerville in the prefrontal and frontal cortex of animals (Gimpl and et al., 2010). Fahrenholz, 2001; Smeltzer et al., 2006), and OTR are found in Based on the above, we think it is likely that chronic OT human cingulate cortex (Boccia et al., 2013). Human imaging modulates human BNST function, and suggest that investi- studies have demonstrated OT-induced modulation of sev- gation of the role of OT in BNST (perhaps with subchronic or eral prefrontal structures in both normals and patients with chronic delivery, utilizing paradigms simulating prolonged anxiety disorders (Bethlehem et al., 2013; Zink and Meyer- anxiety states, and/or in patients who suffer with prolonged Lindenberg, 2012). Specifically, Labuschane performed func- anxiety states (i.e. GAD)) may be illuminating. tional imaging on 18 male social anxiety patients in the context of a face matching task and showed that in addition 4.7.5. Insula to reducing amygdala activation (Labuschagne et al., 2010), Insula, especially anterior insula (AI), is part of the brain's OT normalized hyperactive mPFC and ACC responses to sad salience network (SN) (Seeley et al., 2007). In conjunction with faces (Labuschagne et al., 2011). In a second study examining amygdala, with which it has significant connections, insula resting-state effects of OT in normals, Sripada showed that plays a role in ascertaining the significance of anticipated OT improved connectivity between amygdala and medial stimuli (Lovero et al., 2009; Paulus and Stein, 2006; Stein et al., frontal cortex (Sripada et al., 2012). Together, these studies 2007) and mediating predictive interoceptive signals of aver- invite future studies investigating the role of OT in modulat- sive body states associate with negative outcomes (Alvarez ing activity within prefrontal areas related to anxiety (see et al., 2011; Paulus and Stein, 2006). Hyperactivity of the AI Bethlehem et al. (2013) and Zink and Meyer-Lindenberg (2012) node of the SN has been associated with anxiety disorders for recent reviews). (Etkin and Wager, 2007; Menon, 2011) (but see also Blair et al. In summary: though the study of the effects of OT on (2008)), anxiety sensitivity (hypervigilance for anxiety-related human brain is in its infancy, evidence reviewed above bodily sensations) (Domschke et al., 2010), social anxiety and suggests that OT has effects on several key neural nodes neuroticism (Terasawa et al., 2013), and enhanced autonomic and networks associated with human anxiety disorders, response and startle to threats (Melzig et al., 2008). In regards including PAG, BNST, amygdala, insula, and prefrontal to OT, several recent imaging studies have found OT-induced cortices. modulation of insula and insula-amygdala coupling, though none of these have been examined in connection with anxiety disorders (Domes et al., 2014; Pincus et al., 2010; 5. OT and human anxiety: conclusions and Riem et al., 2012; Rilling, 2013; Striepens et al., 2012). Based on questions this information, examination of the function of insula in anxiety disorder patients treated with OT would be Based on the heterogeneous nature of anxiety disorders and a informative. growing body of data suggesting a complex and nuanced role of the OT system in different aspects of anxiety, we believe it 4.7.6. Prefrontal cortex (ACC, vmPFC, mPFC) is no longer viable to view or describe OT as simply “anxio- A final set of regions associated with anxiety disorders and lytic”. That said, based on the diverse areas of convergent OT are the prefrontal cortices, including ACC, mPFC and research reviewed above, we are optimistic regarding the vmPFC. In general, these regions are discussed in the context potential for OT-targeted treatments for at least some specific of “top-down” modulation of hyperactive or hypersensitive anxiety disorders. Furthermore, we think the diverse litera- subcortical limbic structures and systems (Davidson and tures reviewed above suggest that similarly to the HPA and McEwen, 2012; Kim et al., 2011; Veer et al., 2012). Importantly, serotonin systems, aspects of the OT system will in the future descending influences of the ventromedial prefrontal cortex play a central role in our understanding of the genetics, (vmPFC) and other prefrontal structures on limbic structures epigenetics and development of human anxiety and (i.e. amygdala) have been identified in both rodents and stress-related disorders. Despite the wealth of research humans (Bishop, 2007, Kalin et al., 2007, Peters et al., 2009, above, it is still the case that our understanding of OT as a Soliman et al., 2010). Though significant differences between target for human anxieties is sorely understudied. The field anxiety disorders exists, decreased activation of vmPVC has needs refined translational studies and—most vitally—proof- been shown in a variety of imaging studies of anxiety of-concept randomized clinical trials employing chronic daily patients using different tasks (Killgore et al., 2014; Schienle dosing and meaningful clinical endpoints to help establish et al., 2009; McClure et al., 2007; Krain et al., 2008; Blair et al., OT as a bona-fide monotherapy or adjunctive anxiolytic 2012). Speaking to the effects of medication treatment, a treatment. Coming from a clinically-informed perspective, brain research 1580 (2014) 22–56 43

references Box 1–Questions for future work on OT and anxiety. Acheson, D., Feifel, D., de Wilde, S., McKinney, R., Lohr, J., 1. Can OT or OT mimetics (alone or adjunct to other Risbrough, V., 2013. The effect of intranasal oxytocin treatment medications) significantly reduce symptoms of spe- on conditioned fear extinction and recall in a healthy human cific anxiety disorders? What dose, duration and sample. Psychopharmacology (Berlin) 229 (1), 199–208. delivery system is optimal? What individual factors Adolphs, R., 2008. Fear, faces, and the human amygdala. Curr. Opin. Neurobiol. 18 (2), 166–172. determine response? Are there biomarkers of salu- Adolphs, R., 2009. The social brain: neural basis of social tary effects? knowledge. Annu. Rev. Psychol. 60, 693–716. fi 2. Can OT augment the ef cacy of established psycho- Altemus, M., Jacobson, K.R., Debellis, M., Kling, M., Pigott, T., logical (non-pharmacological) treatments for anxi- Murphy, D.L., et al., 1999. Normal CSF oxytocin and NPY levels ety disorders? in OCD. Biol. Psychiatry 45 (7), 931–933. 3. What role does the OT system play in human fear Alvarez, R.P., Chen, G., Bodurka, J., Kaplan, R., Grillon, C., 2011. learning? How does its role in this process inform Phasic and sustained fear in humans elicits distinct patterns of brain activity. NeuroImage 55 (1), 389–400. the role of OT in the development of different Amico, J.A., Mantella, R.C., Vollmer, R.R., Li, X., 2004. Anxiety and anxiety and anxiety-related disorders? stress responses in female oxytocin deficient mice. J. 4. How does the OT system interact with different Neuroendocrinol. 16 (4), 319–324. neurotransmitters and hormones (i.e. serotonin, 5th ed. American Psychiatric Association, Washington, DC. dopamine, HPA axis) in the genesis and mainte- Amico, J.A., Miedlar, J.A., Cai, H.M., Vollmer, R.R., 2008. nance of anxiety states? Oxytocin knockout mice: a model for studying stress-related 5. What is the role of OT in anxiety symptoms and ingestive behaviours. Prog. Brain Res. 170, 53–64. associated with non-anxiety disorders (i.e. MDD Andari, E., Schneider, F.C., Mottolese, R., Vindras, P., Sirigu, A., with anxious features, eating disorders, schizophre- 2014. Oxytocin’s Fingerprint in Personality Traits and Regional nia, addictions)? Might OT be a helpful adjunctive Brain Volume. Cereb. cortex 24 (2), 479–486. treatment in these conditions? Apter-Levy, Y., Feldman, M., Vakart, A., Ebstein, R.P., Feldman, R., 6. What is the role of the endogenous OT system in 2013. Impact of maternal depression across the first 6 years of common conditions related to anxiety (acute stress life on the child’s mental health, social engagement, and responses, marital conflict, loss of partner)? empathy: the moderating role of oxytocin. Am. J. Psychiatry 170 (10), 1161–1168. 7. How do early experiences (traumatic and protective) Argiolas, A., Melis, M.R., 2013. Neuropeptides and central control canalize the OT system? How do these changes of sexual behaviour from the past to the present: a review. impact the incidence and emergence of anxiety Prog. Neurobiol. 108, 80–107. disorders, including responses to OT? Aupperle, R.L., Ravindran, L., Tankersley, D., Flagan, T., Stein, N.R., 8. How do OT and OT-related genes (i.e. CD38) mod- Simmons, A.N., et al., 2011. Pregabalin influences insula and ulate early experience and protect against or pre- amygdala activation during anticipation of emotional images. dispose to anxiety disorders? Neuropsychopharmacol.: Off. Publ. Am. Coll. Neuropsychopharmacol. 36 (7), 1466–1477. 9. How do individual differences in OT-related genes Ayers, L.W., Missig, G., Schulkin, J., Rosen, J.B., 2011. Oxytocin relate to individual temperamental variables asso- reduces background anxiety in a fear-potentiated startle ciated with anxiety disorders (trait anxiety, neuroti- paradigm: peripheral vs central administration. cism, extraversion, etc.)? How are sex differences in Neuropsychopharmacol.: Off. Publ. Am. Coll. anxiety related? What brain systems mediate these Neuropsychopharmacol. 36 (12), 2488–2497. differences? Baas, J.M., Mol, N., Kenemans, J.L., Prinssen, E.P., Niklson, I., Xia- 10. How do changes in central and peripheral aspects of Chen, C., et al., 2009. Validating a human model for anxiety using startle potentiated by cue and context: the effects of the human OT system (baseline or dynamic OT alprazolam, pregabalin, and diphenhydramine. levels, receptor location or density, OT cell function, Psychopharmacology. (Berl). 205 (1), 73–84. genetic variations in components of the OT system) Baeken, C., De Raedt, R., Ramsey, N., Van Schuerbeek, P., Hermes, relate to different anxiety disorders? D., Bossuyt, A., et al., 2009. Amygdala responses to positively and negatively valenced baby faces in healthy female volunteers: influences of individual differences in harm avoidance. Brain Res. 1296, 94–103. we close with a list of pregnant questions to direct future Bagdy, G., Kalogeras, K.T., 1993. Stimulation of 5-HT1A and 5-HT2/ research in this area (Box 1). 5-HT1C receptors induce oxytocin release in the male rat. Brain Res. 611 (2), 330–332. Bailey, J.E., Argyropoulos, S.V., Kendrick, A.H., Nutt, D.J., 2005.

Behavioral and cardiovascular effects of 7.5% CO2 in human volunteers. Depress. Anxiety 21 (1), 18–25. Acknowledgments Bailey, J.E., Kendrick, A., Diaper, A., Potokar, J.P., Nutt, D.J., 2007. A validation of the 7.5% CO2 model of GAD using paroxetine and lorazepam in healthy volunteers. J. Psychopharmacol. 21 (1), Appreciations are directed to Natalie Ebner, Cort Pedersen, 42–49. Steve Chang, Andres Sciolla, and Tina MacDonald who made Bakermans-Kranenburg, M.J., van Ijzendoorn, M.H., 2013a. helpful suggestions to earlier versions of this MS. Sniffing around oxytocin: review and meta-analyses of trials 44 brain research 1580 (2014) 22–56

in healthy and clinical groups with implications for Bishop, S.J., 2009. Trait anxiety and impoverished prefrontal pharmacotherapy. Transl. Psychiatry 3, e258. control of attention. Nat. Neurosci. 12 (1), 92–98. Bakermans-Kranenburg, M.J., van Ijzendoorn, M.H., 2013b. A Blair, K., Shaywitz, J., Smith, B.W., Rhodes, R., Geraci, M., Jones, sociability gene? Meta-analysis of oxytocin receptor genotype M., et al., 2008. Response to emotional expressions in effects in humans. Psychiatr. Gen. http://dx.doi.org/10.1097/ generalized social phobia and generalized anxiety disorder: YPG.0b013e3283643684. evidence for separate disorders. Am. J. Psychiatry 165 (9), Bales, K.L., Perkeybile, A.M., Conley, O.G., Lee, M.H., Guoynes, C.D., 1193–1202. Downing, G.M., et al., 2013. Chronic intranasal oxytocin causes Blair, K.S., Geraci, M., Smith, B.W., Hollon, N., DeVido, J., long-term impairments in partner preference formation in male Otero, M., et al., 2014. Reduced dorsal anterior cingulate cortical prairie voles. Biol. Psychiatry 74 (3), 180–188. activity during emotional regulation and top-down attentional Bali, B., Kovacs, K.J., 2003. GABAergic control of neuropeptide control in generalized social phobia, generalized anxiety gene expression in parvocellular neurons of the hypothalamic disorder, and comorbid generalized social phobia/generalized paraventricular nucleus. Eur. J. Neurosci. 18 (6), 1518–1526. anxiety disorder. Biol. Psychiatry 72 (6), 476–482. Ball, S., Smolin, J., Shekhar, A., 2002. A psychobiological approach Blair, K.S., Geraci, M., Smith, B.W., Hollon, N., DeVido, J., Otero, M., to personality: examination within anxious outpatients. J. et al., 2012. Reduced dorsal anterior cingulate cortical activity Psychiatric Res. 36 (2), 97–103. during emotional regulation and top-down attentional control Barlow, D.H., 2002. Anxiety and its Disorders: The Nature and in generalized social phobia, generalized anxiety disorder, and Treatment of Anxiety and Panic., 2nd. ed. Guilford Press, comorbid generalized social phobia/generalized anxiety New York. disorder. Biol. Psychiatry. 72 (6), 476–482. Barrett, J., Armony, J.L., 2009. Influence of trait anxiety on brain Blair, R.J., Curran, H.V., 1999. Selective impairment in the activity during the acquisition and extinction of aversive recognition of anger induced by diazepam. conditioning. Psychol. Med. 39 (2), 255–265. Psychopharmacology (Berlin) 147 (3), 335–338. Bartels, A., 2012. Oxytocin and the social brain: beware the Blevins, J.E., Eakin, T.J., Murphy, J.A., Schwartz, M.W., Baskin, D.G., complexity. Neuropsychopharmacol.: Off. Publ. Am. Coll. 2003. Oxytocin innervation of caudal brainstem nuclei Neuropsychopharmacol. 37 (8), 1795–1796. activated by cholecystokinin. Brain Res. 993 (1–2), 30–41. Bartels, A., Zeki, S., 2004. The neural correlates of maternal and Blume, A., Bosch, O.J., Miklos, S., Torner, L., Wales, L., Waldherr, romantic love. Neuroimage 21 (3), 1155–1166. M., et al., 2008. Oxytocin reduces anxiety via ERK1/2 Bartz, J., Simeon, D., Hamilton, H., Kim, S., Crystal, S., Braun, A., et activation: local effect within the rat hypothalamic al., 2011. Oxytocin can hinder trust and cooperation in borderline paraventricular nucleus. Eur. J. Neurosci. 27 (8), personality disorder. Soc. Cogn. Affect. Neurosci. 6 (5), 1947–1956. 556–563. Boccia, M.L., Petrusz, P., Suzuki, K., Marson, L., Pedersen, C.A., Bartz, J.A., 2012. Oxytocin, attachment, betrayal and self-interest: 2013. Immunohistochemical localization of oxytocin receptors a commentary on “oxytocin modulates the link between adult in human brain. Neuroscience 253, 155–164. attachment and cooperation through reduced betrayal den Boer, J.A., Westenberg, H.G., 1992. Oxytocin in obsessive aversion” by Carsten K.W. De Dreu. compulsive disorder. Peptides 13 (6), 1083–1085. Psychoneuroendocrinology. http://dx.doi.org/10.1016/j. Bogels, S.M., Knappe, S., Clark, L.A., 2013. Adult separation psyneuen.2011.10.003. Psychoneuroendocrinology 37 (7), anxiety disorder in DSM-5. Clin. Psychol. Rev. 33 (5), 663–674. 1106–1108; author reply 1108–1110. Bosch, O.J., 2011. Maternal nurturing is dependent on her innate Bearden, C.E., Freimer, N.B., 2006. Endophenotypes for psychiatric anxiety: the behavioral roles of brain oxytocin and disorders: ready for primetime?. Trends Genet.: TIG 22 (6), vasopressin. Horm. Behav. 59 (2), 202–212. 306–313. Bowen, M.T., Carson, D.S., Spiro, A., Arnold, J.C., McGregor, I.S., Beesdo, K., Pine, D.S., Lieb, R., Wittchen, H.U., 2010. Incidence and 2011. Adolescent oxytocin exposure causes persistent risk patterns of anxiety and depressive disorders and reductions in anxiety and alcohol consumption and enhances categorization of generalized anxiety disorder. Arch. Gen. sociability in rats. PLoS ONE 6 (11), e27237. Psychiatry 67 (1), 47–57. Bowlby, J., 1969/1982. Attachment and Loss. vol. 1. Basic Books, Beets, I., Temmerman, L., Janssen, T., Schoofs, L., 2013. Ancient New York. neuromodulation by vasopressin/oxytocin-related peptides. Bradley, B., Davis, T.A., Wingo, A.P., Mercer, K.B., Ressler, K.J., 2013. Worm 2 (2), e24246. Family environment and adult resilience: contributions of positive Belzung, C., Philippot, P., 2007. Anxiety from a phylogenetic parenting and the oxytocin receptor gene. Eur. J. Psychotraumatol. perspective: is there a qualitative difference between human 4, 21659 http://dx.doi.org/10.3402/ejpt.v4i0.21659. and animal anxiety?. Neural. Plast. 2007, 59676. Bradley, B., Westen, D., Mercer, K.B., Binder, E.B., Jovanovic, T., Berns, G.S., Chappelow, J., Cekic, M., Zink, C.F., Pagnoni, G., Crain, D.F., et al., 2011. Association between childhood Martin-Skurski, M.E., 2006. Neurobiological substrates of maltreatment and adult emotional dysregulation in a low- dread. Science 312 (5774), 754–758. income, urban, African American sample: moderation by Bertsch, K., Gamer, M., Schmidt, B., Schmidinger, I., Walther, S., oxytocin receptor gene. Dev. Psychopathol. 23, 439–452. Kastel, T., et al., 2013. Oxytocin and reduction of social threat Brake, W.G., Flores, G., Francis, D., Meaney, M.J., Srivastava, L.K., hypersensitivity in women with borderline personality Gratton, A., 2000. Enhanced nucleus accumbens dopamine disorder. Am. J. Psychiatry. 170 (10), 1169–1177. and plasma corticosterone stress responses in adult rats with Bethlehem, R.A., van Honk, J., Auyeung, B., Baron-Cohen, S., 2013. neonatal excitotoxic lesions to the medial prefrontal cortex. Oxytocin, brain physiology, and functional connectivity: a Neuroscience 96 (4), 687–695. review of intranasal oxytocin fMRI studies. Brennan, B.P., Rauch, S.L., Jensen, J.E., Pope Jr., H.G., 2013. A Psychoneuroendocrinology 38 (7), 962–974. critical review of magnetic resonance spectroscopy studies Biederman, J., Hirshfeld-Becker, D.R., Rosenbaum, J.F., Herot, C., of obsessive-compulsive disorder. Biol. Psychiatry 73 (1), Friedman, D., Snidman, N., et al., 2001. Further evidence of 24–31. association between behavioral inhibition and social anxiety Brennan, K.A., Clark, C.L., Shaver, P.R., 1998. Self-report in children. Am. J. Psychiatry 158 (10), 1673–1679. measurement of adult attachment: an integrative overview. Bishop, S.J., 2007. Neurocognitive mechanisms of anxiety: an In: Simpson, J.A., Rholes, W.S. (Eds.), Attachment Theory and integrative account. Trends Cogn. Sci. 11 (7), 307–316. Close Relationships. Guilford Press, New York, pp. 46–76. brain research 1580 (2014) 22–56 45

Brown, C.H., Bains, J.S., Ludwig, M., Stern, J.E., 2013. Physiological Choi, J.M., Padmala, S., Pessoa, L., 2012. Impact of state anxiety on regulation of magnocellular neurosecretory cell activity: the interaction between threat monitoring and cognition. integration of intrinsic, local and afferent mechanisms. J. NeuroImage 59 (2), 1912–1923. Neuroendocrinol. 25 (8), 678–710. Chrousos, G.P., 1995. The hypothalamic-pituitary-adrenal axis Brune, M., 2012. Does the oxytocin receptor (OXTR) and immune-mediated inflammation. N. Engl. J. Med. 332 (20), polymorphism (rs2254298) confer “vulnerability” for 1351–1362. psychopathology or “differential susceptibility”? Insights from Clark, C.L., St. John, N., Pasca, A.M., Hyde, S.A., Hornbeak, K., evolution. BMC. Med. 10 (1), 38. Abramova, M., et al., 2013. Neonatal CSF oxytocin levels are Brune, M., Brune-Cohrs, U., 2006. Theory of mind—evolution, associated with parent report of infant soothability and ontogeny, brain mechanisms and psychopathology. Neurosci. sociability. Psychoneuroendocrinology 38 (7), 1208–1212. Biobehav. Rev. 30 (4), 437–455. Clark, D.A., Beck, A.T., Beck, J.S., 1994. Symptom differences in Bryant, R.A., Felmingham, K., Kemp, A., Das, P., Hughes, G., Peduto, major depression, dysthymia, panic disorder, and generalized A., et al., 2008. Amygdala and ventral anterior cingulate activation anxiety disorder. Am. J. Psychiatry 151 (2), 205–209. predicts treatment response to cognitive behaviour therapy Cloninger, C.R., 1987. A systematic method for clinical for post-traumatic stress disorder. Psychol. Med. 38 (4), description and classification of personality variants. A 555–561. proposal. Arch. Gen. Psychiatry 44 (6), 573–588. Buhle, J.T., Kober, H., Ochsner, K.N., Mende-Siedlecki, P., Weber, J., Corr, P.J., Kumari, V., Wilson, G.D., Checkley, S., Gray, J.A., 1997. Hughes, B.L., et al., 2013. Common representation of pain and Harm avoidance and affective modulation of the startle reflex: negative emotion in the midbrain periaqueductal gray. Soc. a replication. Pers. Individ. Differ. 22, 591–593. Cogn. Affect. Neurosci. 8 (6), 609–616. Costa, B., Pini, S., Gabelloni, P., Abelli, M., Lari, L., Cardini, A., et Buijs, R.M., Velis, D.N., Swaab, D.F., 1980. Extrahypothalamic al., 2009. Oxytocin receptor polymorphisms and adult vasopressin and oxytocin innervation of fetal and adult rat attachment style in patients with depression. brain. Prog. Brain Res. 53, 159–167. Psychoneuroendocrinology. 34 (10), 1506–1514. Burghardt, N.S., Bauer, E.P., 2013. Acute and chronic effects of Coupland, N.J., Singh, A.J., Sustrik, R.A., Ting, P., Blair, R., 2003. selective serotonin reuptake inhibitor treatment on fear Effects of diazepam on facial emotion recognition. J. conditioning: Implications for underlying fear circuits. Psychiatry Neurosci.: JPN 28 (6), 452–463. Neuroscience 247, 253–272. Craske, M.G., Roy-Byrne, P.P., Stein, M.B., Sullivan, G., Sherbourne, Burghy, C.A., Stodola, D.E., Ruttle, P.L., Molloy, E.K., Armstrong, J. C., Bystritsky, A., 2009. Treatment for anxiety disorders: M., Oler, J.A., et al., 2012. Developmental pathways to efficacy to effectiveness to implementation. Behav. Res. Ther. amygdala-prefrontal function and internalizing symptoms in 47 (11), 931–937. adolescence. Nat. Neurosci. 15 (12), 1736–1741. Craske, M.G., Rauch, S.L., Ursano, R., Prenoveau, J., Pine, D.S., Burri, A., Heinrichs, M., Schedlowski, M., Kruger, T.H., 2008. The Zinbarg, R.E., 2009. What is an anxiety disorder?. Depress. acute effects of intranasal oxytocin administration on Anxiety 26 (12), 1066–1085. endocrine and sexual function in males. Csete, M., Doyle, J., 2004. Bow ties, metabolism and disease. Psychoneuroendocrinology 33 (5), 591–600. Trends Biotechnol. 22 (9), 446–450. Campbell-Sills, L., Simmons, A.N., Lovero, K.L., Rochlin, A.A., Csete, M.E., Doyle, J.C., 2002. Reverse engineering of biological Paulus, M.P., Stein, M.B., 2011. Functioning of neural systems complexity. Science 295 (5560), 1664–1669. supporting emotion regulation in anxiety-prone individuals. D’Avanzato, C., Martinez, J., Attiullah, N., Friedman, M., Toba, C., NeuroImage 54 (1), 689–696. Boerescu, D.A., et al., 2013. Anxiety symptoms among Cannon, W.B., 1932. The Wisdom of the Body. W. W. Norton & remitted depressed outpatients: prevalence and association Company, Inc, New York. with quality of life and psychosocial functioning. J. Affect. Cardoso, C., Ellenbogen, M.A., Orlando, M.A., Bacon, S.L., Joober, Disord. 151 (1), 401–404. R., 2013. Intranasal oxytocin attenuates the cortisol response Dabrowska, J., Hazra, R., Ahern, T.H., Guo, J.D., McDonald, A.J., to physical stress: a dose–response study. Mascagni, F., et al., 2011. Neuroanatomical evidence for Psychoneuroendocrinology 38 (3), 399–407. reciprocal regulation of the corticotrophin-releasing factor Carter, C.S., Krug, M.K., 2009. The functional neuroanatomy of and oxytocin systems in the hypothalamus and the bed dread: Functional magnetic resonance imaging insights into nucleus of the stria terminalis of the rat: implications for generalized anxiety disorder and its treatment. Am. J balancing stress and affect. Psychoneuroendocrinology 36 (9), Psychiatry. 166 (3), 263–265. 1312–1326. Carrasco, G.A., Van de Kar, L.D., 2003. Neuroendocrine Dabrowska, J., Hazra, R., Guo, J.D., Dewitt, S., Rainnie, D.G., 2013. pharmacology of stress. Eur. J. Pharmacol. 463 (1–3), Central CRF neurons are not created equal: phenotypic 235–272. differences in CRF-containing neurons of the rat Carson, W.H., Kitagawa, H., Nemeroff, C.B., 2004. Drug paraventricular hypothalamus and the bed nucleus of the development for anxiety disorders: new roles for atypical stria terminalis. Front. Neurosci. 7, 156. antipsychotics. Psychopharmacol. Bull. 38 (Suppl. 1), S38–S45. Damasio, A.R., Grabowski, T.J., Bechara, A., Damasio, H., Ponto, L. Champagne, F., Meaney, M.J., 2001. Like mother, like daughter: L., Parvizi, J., et al., 2000. Subcortical and cortical brain activity evidence for non-genomic transmission of parental behavior during the feeling of self-generated emotions. Nat. Neurosci. 3 and stress responsivity. Prog. Brain Res. 133, 287–302. (10), 1049–1056. Champagne, F., Diorio, J., Sharma, S., Meaney, M.J., 2001. Naturally Damsa, C., Kosel, M., Moussally, J., 2009. Current status of brain occurring variations in maternal behavior in the rat are imaging in anxiety disorders. Curr. Opin. Psychiatry 22 (1), associated with differences in estrogen-inducible central 96–110. oxytocin receptors. Proc. Natl. Acad. Sci. USA 98 (22), 12736–12741. Dannlowski, U., Stuhrmann, A., Beutelmann, V., Zwanzger, P., Champagne, F.A., 2008. Epigenetic mechanisms and the Lenzen, T., Grotegerd, D., et al., 2012. Limbic scars: long-term transgenerational effects of maternal care. Front. consequences of childhood maltreatment revealed by Neuroendocrinol. 29 (3), 386–397. functional and structural magnetic resonance imaging. Biol. Chang, S.W., Brent, L.J., Adams, G.K., Klein, J.T., Pearson, J.M., Psychiatry 71 (4), 286–293. Watson, K.K., et al., 2013. Neuroethology of primate social Dannlowski, U., Kugel, H., Huber, F., Stuhrmann, A., Redlich, R., behavior. Proc. Natl. Acad. Sci. USA 110 (Suppl. 2), S10387–S10394. Grotegerd, D., et al., 2013. Childhood maltreatment is 46 brain research 1580 (2014) 22–56

associated with an automatic negative emotion processing Domschke, K., Stevens, S., Pfleiderer, B., Gerlach, A.L., 2010. bias in the amygdala. Hum. Brain Mapp. 34 (11), 2899–2909. Interoceptive sensitivity in anxiety and anxiety disorders: an Davidson, J.R., Feltner, D.E., Dugar, A., 2010. Management of overview and integration of neurobiological findings. Clin. generalized anxiety disorder in primary care: identifying the Psychol. Rev. 30 (1), 1–11. challenges and unmet needs. Prim. Care Companion J. Clin. Dunbar, R., 2003. Psychology. Evolution of the social brain. Psychiatry 12, 2. Science 302 (5648), 1160–1161. Davidson, R.J., McEwen, B.S., 2012. Social influences on Eaton, J.L., Roache, L., Nguyen, K.N., Cushing, B.S., Troyer, E., neuroplasticity: stress and interventions to promote well- Papademetriou, E., et al., 2012. Organizational effects of oxytocin being. Nat. Neurosci. 15 (5), 689–695. on serotonin innervation. Dev. Psychobiol. 54 (1), 92–97. Davis, M., Walker, D.L., Miles, L., Grillon, C., 2010. Phasic vs Ehringer, M.A., Rhee, S.H., Young, S., Corley, R., Hewitt, J.K., 2006. sustained fear in rats and humans: role of the extended Genetic and environmental contributions to common amygdala in fear vs. anxiety. Neuropsychopharmacol.: Off. psychopathologies of childhood and adolescence: a study of Publ. Am. Coll. Neuropsychopharmacol. 35 (1), 105–135. twins and their siblings. J. Abnorm. Child Psychol. 34 (1), 1–17. De Dreu, C.K., 2012. Oxytocin, attachment, and self-regarding Elton, A., Tripathi, S.P., Mletzko, T., Young, J., Cisler, J.M., James, G. preferences in humans: rejoinder to Bartz. A., et al., 2013. Childhood maltreatment is associated with a Psychoneuroendocrinologyhttp://dxdoi.org/10.1016/j. sex-dependent functional reorganization of a brain inhibitory psyneuen.2012.1003.1004. control network. Hum. Brain Mapp. (first published online, De Dreu, C.K., Greer, L.L., Van Kleef, G.A., Shalvi, S., Handgraaf, M. http://dx.doi.org/10.1002/hbm.22280). J., 2011. Oxytocin promotes human ethnocentrism. Proc. Natl. Emery, N.J., 2000. The eyes have it: the neuroethology, function Acad. Sci. USA 108 (4), 1262–1266. and evolution of social gaze. Neurosci. Biobehav. Rev. 24 (6), Deakin, J.F., Graeff, F.G., 1991. 5-HT and mechanisms of defence. J. 581–604. Psychopharmacol. 5 (4), 305–315. Engelmann, M., Landgraf, R., Wotjak, C.T., 2004. The Del-Ben, C.M., Deakin, J.F., McKie, S., Delvai, N.A., Williams, S.R., hypothalamic-neurohypophysial system regulates the Elliott, R., et al., 2005. The effect of citalopram pretreatment hypothalamic-pituitary-adrenal axis under stress: an old on neuronal responses to neuropsychological tasks in normal concept revisited. Front. Neuroendocrinol. 25 (3–4), 132–149. volunteers: an FMRI study. Neuropsychopharmacol.: Off. Publ. Epperson, C.N., McDougle, C.J., Price, L.H., 1996. Intranasal Am. Coll. Neuropsychopharmacol. 30 (9), 1724–1734. oxytocin in obsessive-compulsive disorder. Biol. Psychiatry 40 Del-Ben, C.M., Ferreira, C.A., Sanchez, T.A., Alves-Neto, W.C., (6), 547–549. Guapo, V.G., de Araujo, D.B., et al., 2012. Effects of diazepam on Esbjorn, B.H., Bender, P.K., Reinholdt-Dunne, M.L., Munck, L.A., BOLD activation during the processing of aversive faces. J. Ollendick, T.H., 2012. The development of anxiety disorders: Psychopharmacol. 26 (4), 443–451. considering the contributions of attachment and emotion de Oliveira, D.C., Chagas, M.H., Garcia, L.V., Crippa, J.A., Zuardi, A. regulation. Clin. Child Fam. Psychol. Rev. 15 (2), 129–143. W., 2012a. Oxytocin interference in the effects induced by Etkin, A., Wager, T.D., 2007. Functional neuroimaging of anxiety: a inhalation of 7.5% CO(2) in healthy volunteers. Hum. meta-analysis of emotional processing in PTSD, social anxiety Psychopharmacol. 27 (4), 378–385. disorder, and specific phobia. Am. J. Psychiatry 164 (10), de Oliveira, D.C., Zuardi, A.W., Graeff, F.G., Queiroz, R.H., 1476–1488. Crippa, J.A., 2012b. Anxiolytic-like effect of oxytocin in the Etkin, A., Prater, K.E., Schatzberg, A.F., Menon, V., Greicius, M.D., simulated public speaking test. J. Psychopharmacol. 26 (4), 2009. Disrupted amygdalar subregion functional 497–504. connectivity and evidence of a compensatory network in Dickerson, S.S., Kemeny, M.E., 2004. Acute stressors and cortisol generalized anxiety disorder. Arch. Gen. Psychiatry 66 (12), responses: a theoretical integration and synthesis of 1361–1372. laboratory research. Psychol. Bull. 130 (3), 355–391. Etkin, A., Prater, K.E., Hoeft, F., Menon, V., Schatzberg, A.F., 2010. Diorio, D., Viau, V., Meaney, M.J., 1993. The role of the medial Failure of anterior cingulate activation and connectivity with prefrontal cortex (cingulate gyrus) in the regulation of the amygdala during implicit regulation of emotional hypothalamic-pituitary-adrenal responses to stress. J. processing in generalized anxiety disorder. Am. J. Psychiatry Neurosci.: Off. J. Soc. Neurosci. 13 (9), 3839–3847. 167 (5), 545–554. Ditzen, B., Schaer, M., Gabriel, B., Bodenmann, G., Ehlert, U., Farmer, A., Mahmood, A., Redman, K., Harris, T., Sadler, S., Heinrichs, M., 2009. Intranasal oxytocin increases positive McGuffin, P., 2003. A sib-pair study of the temperament and communication and reduces cortisol levels during couple character inventory scales in major depression. Arch. Gen. conflict. Biol. Psychiatry 65 (9), 728–731. Psychiatry 60 (5), 490–496. Dolen, G., Darvishzadeh, A., Huang, K.W., Malenka, R.C., 2013. Fava, M., Rush, A.J., Alpert, J.E., Balasubramani, G.K., Wisniewski, Social reward requires coordinated activity of nucleus S.R., Carmin, C.N., et al., 2008. Difference in treatment accumbens oxytocin and serotonin. Nature 501 (7466), outcome in outpatients with anxious versus nonanxious 179–184. depression: a STARnD report. Am. J. Psychiatry 165 (3), Domes, G., Heinrichs, M., Kumbier, E., Grossmann, A., 342–351. Hauenstein, K., Herpertz, S.C., 2013a. Effects of intranasal Feifel, D., Reza, T., 1999. Oxytocin modulates psychotomimetic- oxytocin on the neural basis of face processing in autism induced deficits in sensorimotor gating. Psychopharmacology spectrum disorder. Biol. Psychiatry 74 (3), 164–171. (Berlin) 141 (1), 93–98. Domes, G., Steiner, A., Porges, S.W., Heinrichs, M., 2013b. Feifel, D., MacDonald, K., McKinney, R., Heisserer, N., Serrano, V., Oxytocin differentially modulates eye gaze to naturalistic 2011. A randomized, placebo-controlled investigation of social signals of happiness and anger. intranasal oxytocin in patients with anxiety. Psychoneuroendocrinology 38 (7), 1198–1202. Neuropsychopharmacology 36, S324–S449. Domes, G., Kumbier, E., Heinrichs, M., & Herpertz, S. C. (2014). Feifel, D., Shilling, P.D., Belcher, A.M., 2012. The effects of oxytocin Oxytocin promotes facial emotion recognition and amygdala and its analog, carbetocin, on genetic deficits in sensorimotor reactivity in adults with asperger syndrome. gating. Eur. Neuropsychopharmacol.: J. Eur. Coll. Neuropsychopharmacology : official publication of the Neuropsychopharmacol. 22 (5), 374–378. American College of Neuropsychopharmacology, 39(3), Feldman, R., 2007. Parent-infant synchrony and the construction 698-706. of shared timing; physiological precursors, developmental brain research 1580 (2014) 22–56 47

outcomes, and risk conditions. J. Child Psychol. Psychiatry 48 healthy volunteers: a fear-potentiated startle study. (3–4), 329–354. Neuropsychopharmacology. 34 (4), 964–971. Feldman, R., 2012. Oxytocin and social affiliation in humans. Grippo, A.J., Trahanas, D.M., Zimmerman 2nd, R.R., Porges, S.W., Horm. Behav. 61 (3), 380–391. Carter, C.S., 2009. Oxytocin protects against negative Feldman, R., Gordon, I., Zagoory-Sharon, O., 2010. The cross- behavioral and autonomic consequences of long-term generation transmission of oxytocin in humans. Horm. Behav. social isolation. Psychoneuroendocrinology 34 (10), 58 (4), 669–676. 1542–1553. Feldman, R., Gordon, I., Influs, M., Gutbir, T., Ebstein, R.P., 2013. Grippo, A.J., Pournajafi-Nazarloo, H., Sanzenbacher, L., Trahanas, Parental oxytocin and early caregiving jointly shape children’s D.M., McNeal, N., Clarke, D.A., et al., 2012. Peripheral oxytocin oxytocin response and social reciprocity. administration buffers autonomic but not behavioral Neuropsychopharmacol.: Off. Publ. Am. Coll. responses to environmental stressors in isolated prairie voles. Neuropsychopharmacol. 38 (7), 1154–1162. Stress 15 (2), 149–161. Fernando, A.B., Robbins, T.W., 2011. Animal models of Grillon, C., Krimsky, M., Charney, D.R., Vytal, K., Ernst, M., neuropsychiatric disorders. Annu. Rev. Clin. Psychol. 7, 39–61. Cornwell, B., 2013. Oxytocin increases anxiety to Figueira, R.J., Peabody, M.F., Lonstein, J.S., 2008. Oxytocin receptor unpredictable threat. Mol. Psychiatry. 18 (9), 958–960. activity in the ventrocaudal periaqueductal gray modulates Groppe, S.E., Gossen, A., Rademacher, L., Hahn, A., Westphal, L., anxiety-related behavior in postpartum rats. Behav. Neurosci. Grunder, G., et al., 2013. Oxytocin influences processing of 122 (3), 618–628. socially relevant cues in the ventral tegmental area of the Foa, E.B., 2006. Psychosocial therapy for posttraumatic stress human brain. Biol. Psychiatry 74 (3), 172–179. disorder. J. Clin. Psychiatry 67 (Suppl. 2), S40–S45. Guastella, A.J., Howard, A.L., Dadds, M.R., Mitchell, P., Carson, D. Fox, A.S., Shelton, S.E., Oakes, T.R., Davidson, R.J., Kalin, N.H., S., 2009. A randomized controlled trial of intranasal oxytocin 2008. Trait-like brain activity during adolescence predicts as an adjunct to exposure therapy for social anxiety disorder. anxious temperament in primates. PLoS ONE 3 (7), e2570. Psychoneuroendocrinology 34 (6), 917–923. Fox, N.A., Henderson, H.A., Marshall, P.J., Nichols, K.E., Ghera, M.M., Guzman, Y.F., Tronson, N.C., Jovasevic, V., Sato, K., Guedea, A.L., 2005. Behavioral inhibition: linking biology and behavior Mizukami, H., et al., 2013. Fear-enhancing effects of septal within a developmental framework. Annu. Rev. Psychol. 56, oxytocin receptors. Nature neuroscience 16 (9), 1185–1187. 235–262. Hall, S.S., Lightbody, A.A., McCarthy, B.E., Parker, K.J., Reiss, A.L., 2012. Gamer, M., Zurowski, B., Buchel, C., 2010. Different amygdala Effects of intranasal oxytocin on social anxiety in males with subregions mediate valence-related and attentional effects of fragile X syndrome. Psychoneuroendocrinology 37 (4), 509–518. oxytocin in humans. Proc. Natl. Acad. Sci. USA 107 (20), Hammack, S.E., Richey, K.J., Watkins, L.R., Maier, S.F., 2004. 9400–9405. Chemical lesion of the bed nucleus of the stria terminalis Garrison, J.L., Macosko, E.Z., Bernstein, S., Pokala, N., Albrecht, D. blocks the behavioral consequences of uncontrollable stress. R., Bargmann, C.I., 2012. Oxytocin/vasopressin-related Behav. Neurosci. 118 (2), 443–448. peptides have an ancient role in reproductive behavior. Heim, C., Nemeroff, C.B., 2001. The role of childhood trauma in Science 338 (6106), 540–543. the neurobiology of mood and anxiety disorders: preclinical Ge, Y., Lundeberg, T., Yu, L.C., 2002. Blockade effect of mu and and clinical studies. Biol. Psychiatry 49 (12), 1023–1039. kappa opioid antagonists on the anti-nociception induced by Heim, C., Newport, D.J., Mletzko, T., Miller, A.H., Nemeroff, C.B., intra-periaqueductal grey injection of oxytocin in rats. Brain 2008. The link between childhood trauma and depression: Res. 927 (2), 204–207. insights from HPA axis studies in humans. Gee, D.G., Gabard-Durnam, L.J., Flannery, J., Goff, B., Humphreys, Psychoneuroendocrinology 33 (6), 693–710. K.L., Telzer, E.H., et al., 2013. Early developmental emergence Heim, C., Young, L.J., Newport, D.J., Mletzko, T., Miller, A.H., of human amygdala-prefrontal connectivity after maternal Nemeroff, C.B., 2009. Lower CSF oxytocin concentrations in deprivation. Proc. Natl. Acad. Sci. USA 110 (39), 15638–15643. women with a history of childhood abuse. Mol. Psychiatry 14 Gimpl, G., Fahrenholz, F., 2001. The oxytocin receptor system: (10), 954–958. structure, function, and regulation. Physiol. Rev. 81 (2), Heim, C., Shugart, M., Craighead, W.E., Nemeroff, C.B., 2010. 629–683. Neurobiological and psychiatric consequences of child abuse Goodson, J.L., 2013. Deconstructing sociality, social evolution and and neglect. Dev. Psychobiol. 52 (7), 671–690. relevant nonapeptide functions. Psychoneuroendocrinology Heinrichs, M., Baumgartner, T., Kirschbaum, C., Ehlert, U., 2003. 38 (4), 465–478. Social support and oxytocin interact to suppress cortisol and Gottesman, II, Gould, T.D, 2003. The endophenotype concept in subjective responses to psychosocial stress. Biol. Psychiatry 54 psychiatry: etymology and strategic intentions. Am. J. (12), 1389–1398. Psychiatry 160 (4), 636–645. Hek, K., Direk, N., Newson, R.S., Hofman, A., Hoogendijk, W.J., Grant, M.M., Cannistraci, C., Hollon, S.D., Gore, J., Shelton, R., Mulder, C.L., et al., 2013. Anxiety disorders and salivary 2011. Childhood trauma history differentiates amygdala cortisol levels in older adults: a population-based study. response to sad faces within MDD. J. Psychiatr. Res. 45 (7), Psychoneuroendocrinology 38 (2), 300–305. 886–895. Hellhammer, D.H., Wust, S., Kudielka, B.M., 2009. Salivary cortisol Grewen, K.M., Girdler, S.S., Amico, J., Light, K.C., 2005. Effects of as a biomarker in stress research. Psychoneuroendocrinology partner support on resting oxytocin, cortisol, norepinephrine, 34 (2), 163–171. and blood pressure before and after warm partner contact. Herman, J.P., Cullinan, W.E., 1997. Neurocircuitry of stress: central Psychosom. Med. 67 (4), 531–538. control of the hypothalamo-pituitary-adrenocortical axis. Grillon, C., Lissek, S., Rabin, S., McDowell, D., Dvir, S., Pine, D.S., Trends Neurosci. 20 (2), 78–84. 2008. Increased anxiety during anticipation of unpredictable Hermans, E.J., Henckens, M.J., Roelofs, K., Fernandez, G., 2012. but not predictable aversive stimuli as a psychophysiologic Fear bradycardia and activation of the human periaqueductal marker of panic disorder. Am. J. Psychiatry 165 (7), grey. NeuroImage 66C, 278–287. 898–904. Hesketh, S., Jessop, D.S., Hogg, S., Harbuz, M.S., 2005. Differential Grillon, C., Chavis, C., Covington, M.F., Pine, D.S., 2009. Two-week actions of acute and chronic citalopram on the rodent treatment with the selective serotonin reuptake inhibitor hypothalamic-pituitary-adrenal axis response to acute citalopram reduces contextual anxiety but not cued fear in restraint stress. J. Endocrinol. 185 (3), 373–382. 48 brain research 1580 (2014) 22–56

Hettema, J.M., An, S.S., Bukszar, J., van den Oord, E.J., Neale, M.C., and amygdalar volume in healthy adults. Biol. Psychiatry 68 Kendler, K.S., et al., 2008. Catechol-O-methyltransferase (11), 1066–1072. contributes to genetic susceptibility shared among anxiety Inoue, H., Yamasue, H., Tochigi, M., Abe, O., Liu, X., Kawamura, Y., spectrum phenotypes. Biol. Psychiatry 64 (4), 302–310. et al., 2011. Association between the oxytocin receptor gene Hicks, C., Jorgensen, W., Brown, C., Fardell, J., Koehbach, J., and amygdalar volume in healthy adults. Biol. Psychiatry 68 Gruber, C.W., et al. (2012). The non-peptide oxytocin receptor (11), 1066–1072. agonist WAY 267,464: receptor binding profile, prosocial Insel, T., Cuthbert, B., Garvey, M., Heinssen, R., Pine, D.S., Quinn, effects and distribution of c-Fos expression in adolescent rats. K., et al., 2010. Research domain criteria (RDoC): toward a new J. Neuroendocrinol.. classification framework for research on mental disorders. Hirshfeld, D.R., Rosenbaum, J.F., Biederman, J., Bolduc, E.A., Am. J. Psychiatry 167 (7), 748–751. Faraone, S.V., Snidman, N., et al., 1992. Stable behavioral Insel, T.R., 1992. Oxytocin—a neuropeptide for affiliation: inhibition and its association with anxiety disorder. J. Am. evidence from behavioral, receptor autoradiographic, and Acad. Child Adolesc. Psychiatry 31 (1), 103–111. comparative studies. Psychoneuroendocrinology 17 (1), Hirshfeld-Becker, D.R., Micco, J.A., Simoes, N.A., Henin, A., 2008. 3–35. High risk studies and developmental antecedents of anxiety Insel, T.R., 1997. A neurobiological basis of social attachment. disorders. Am. J. Med. Genet. C. Semin. Med. Genet. 148C (2), Am. J. Psychiatry 154 (6), 726–735. 99–117. Insel, T.R., 2003. Is social attachment an addictive disorder?. Hodgson, R., Rachman, S., 1974. II. Desynchrony in measures of Physiol. Behav. 79 (3), 351–357. fear. Behav. Res. Ther. 12 (4), 319–326. Jack, A., Connelly, J.J., Morris, J.P., 2012. DNA methylation of the Hofer, M.A., 1994a. Early relationships as regulators of infant oxytocin receptor gene predicts neural response to ambiguous physiology and behavior. Acta. Paediatr. Suppl. 397, 9–18. social stimuli. Front. Hum. Neurosci. 6, 280. Hofer, M.A., 1994b. Hidden regulators in attachment, separation, Jenkins, J.S., Ang, V.T., Hawthorn, J., Rossor, M.N., Iversen, L.L., and loss. Monogr. Soc. Res. Child Dev. 59 (2–3), 192–207. 1984. Vasopressin, oxytocin and neurophysins in the human Hoge, E.A., Pollack, M.H., Kaufman, R.E., Zak, P.J., Simon, N.M., brain and spinal cord. Brain Res. 291 (1), 111–117. 2008. Oxytocin levels in social anxiety disorder. CNS. Neurosci. Jin, D., Liu, H.X., Hirai, H., Torashima, T., Nagai, T., Lopatina, O., et Ther. 14 (3), 165–170. al., 2007. CD38 is critical for social behaviour by regulating Hoge, E.A., Lawson, E.A., Metcalf, C.A., Keshaviah, A., Zak, P.J., oxytocin secretion. Nature 446 (7131), 41–45. Pollack, M.H., et al., 2012. Plasma oxytocin immunoreactive Jokinen, J., Chatzittofis, A., Hellstrom, C., Nordstrom, P., Uvnas- products and response to trust in patients with social anxiety Moberg, K., Asberg, M., 2012. Low CSF oxytocin reflects high disorder. Depress. Anxiety 29 (11), 924–930. intent in suicide attempters. Psychoneuroendocrinology 37 Hollander, E., Novotny, S., Hanratty, M., Yaffe, R., DeCaria, C.M., (4), 482–490. Aronowitz, B.R., et al., 2003. Oxytocin infusion reduces Jonas, W., Mileva-Seitz, V., Girard, A.W., Bisceglia, R., Kennedy, J.L., repetitive behaviors in adults with autistic and Sokolowski, M., et al., 2013. Genetic variation in oxytocin Asperger’s disorders. Neuropsychopharmacology 28 (1), rs2740210 and early adversity associated with postpartum 193–198. depression and breastfeeding duration. Genes Brain Behav. 12 Honey, C.J., Kotter, R., Breakspear, M., Sporns, O., 2007. Network (7), 681–694. structure of cerebral cortex shapes functional connectivity on Kagan, J., 2008. Using the proper vocabulary. Dev. Psychobiol. 50 multiple time scales. Proc. Natl. Acad. Sci. USA 104 (24), (1), 4–8. 10240–10245. Kagan, J., Snidman, N., McManis, M., Woodward, S., 2001. Honey, C.J., Sporns, O., Cammoun, L., Gigandet, X., Thiran, J.P., Temperamental contributions to the affect family of anxiety. Meuli, R., et al., 2009. Predicting human resting-state Psychiatr. Clin. N. Am. 24 (4), 677–688. functional connectivity from structural connectivity. Proc. Kagerbauer, S.M., Martin, J., Schuster, T., Blobner, M., Kochs, E.F., Natl. Acad. Sci. USA 106 (6), 2035–2040. Landgraf, R., 2013. Plasma oxytocin and vasopressin do not van Honk, J., Peper, J.S., Schutter, D.J., 2005. Testosterone reduces predict neuropeptide concentrations in human cerebrospinal unconscious fear but not consciously experienced anxiety: fluid. J. Neuroendocrinol. 25 (7), 668–673. implications for the disorders of fear and anxiety. Biol. Kalin, N.H., Shelton, S.E., Davidson, R.J., 2007. Role of the primate Psychiatry 58 (3), 218–225. orbitofrontal cortex in mediating anxious temperament. Biol. Hoyda, T.D., Ferguson, A.V., 2010. Adiponectin modulates Psychiatry. 62 (10), 1134–1139. excitability of rat paraventricular nucleus neurons by Kamboj, S.K., Curran, H.V., 2006. Scopolamine induces differential modulation of potassium currents. Endocrinology impairments in the recognition of human facial expressions 151 (7), 3154–3162. of anger and disgust. Psychopharmacology (Berlin) 185 (4), Hrabovszky, E., Liposits, Z., 2008. Novel aspects of glutamatergic 529–535. signalling in the neuroendocrine system. J. Neuroendocrinol. Keebaugh, A.C., Young, L.J., 2011. Increasing oxytocin receptor 20 (6), 743–751. expression in the nucleus accumbens of pre-pubertal female Huang, H., Michetti, C., Busnelli, M., Manago, F., Sannino, S., prairie voles enhances alloparental responsiveness and Scheggia, D., et al., 2013. Chronic and acute intranasal oxytocin partner preference formation as adults. Horm. Behav. 60 (5), produce divergent social effects in mice. 498–504. Neuropsychopharmacol.: Off. Publ. Am. Coll. Kemp, A.H., Gray, M.A., Silberstein, R.B., Armstrong, S.M., Nathan, Neuropsychopharmacol. http://dx.doi.org/10.1038/npp.2013.310. P.J., 2004. Augmentation of serotonin enhances pleasant and Huber, D., Veinante, P., Stoop, R., 2005. Vasopressin and oxytocin suppresses unpleasant cortical electrophysiological responses excite distinct neuronal populations in the central amygdala. to visual emotional stimuli in humans. NeuroImage 22 (3), Science 308 (5719), 245–248. 1084–1096. Indovina, I., Robbins, T.W., Nunez-Elizalde, A.O., Dunn, B.D., Kendler, K.S., 2005. Toward a philosophical structure for Bishop, S.J., 2011. Fear-conditioning mechanisms associated psychiatry. Am. J. Psychiatry 162 (3), 433–440. with trait vulnerability to anxiety in humans. Neuron 69 (3), Kendler, K.S., 2008. Explanatory models for psychiatric illness. 563–571. Am. J. Psychiatry 165 (6), 695–702. Inoue, H., Yamasue, H., Tochigi, M., Abe, O., Liu, X., Kawamura, Y., Kendler, K.S., Prescott, C.A., Myers, J., Neale, M.C., 2003. The et al., 2010. Association between the oxytocin receptor gene structure of genetic and environmental risk factors for brain research 1580 (2014) 22–56 49

common psychiatric and substance use disorders in men and Koolhaas, J.M., Bartolomucci, A., Buwalda, B., de Boer, S.F., Flugge, women. Arch. Gen. Psychiatry 60 (9), 929–937. G., Korte, S.M., et al., 2011. Stress revisited: a critical evaluation Kendler, K.S., Kuhn, J.W., Prescott, C.A., 2004. Childhood sexual of the stress concept. Neurosci. Biobehav. Rev. 35 (5), 1291–1301. abuse, stressful life events and risk for major depression in Krain, A.L., Gotimer, K., Hefton, S., Ernst, M., Castellanos, F.X., women. Psychol. Med. 34 (8), 1475–1482. Pine, D.S., et al., 2008. A functional magnetic resonance Kendler, K.S., Aggen, S.H., Knudsen, G.P., Roysamb, E., imaging investigation of uncertainty in adolescents with Neale, M.C., Reichborn-Kjennerud, T., 2011. The structure of anxiety disorders. Biol. Psychiatry. 63 (6), 563–568. genetic and environmental risk factors for syndromal and Krueger, R.F., 1999. The structure of common mental disorders. subsyndromal common DSM-IV axis I and all axis II disorders. Arch. Gen. Psychiatry 56 (10), 921–926. Am. J. Psychiatry 168 (1), 29–39. Kumsta, R., Heinrichs, M., 2013. Oxytocin, stress and social Kenkel, W.M., Paredes, J., Lewis, G.F., Yee, J.R., Pournajafi-Nazarloo, behavior: neurogenetics of the human oxytocin system. Curr. H., Grippo, A.J., et al., 2013. Autonomic substrates of the Opin. Neurobiol. 23 (1), 11–16. response to pups in male prairie voles. PloS One 8 (8), e69965. Kumsta, R., Hummel, E., Chen, F.S., Heinrichs, M., 2013. Epigenetic Keri, S., Kiss, I., Kelemen, O., 2009. Sharing secrets: oxytocin and regulation of the oxytocin receptor gene: implications for trust in schizophrenia. Soc. Neurosci. 4 (4), 287–293. behavioral neuroscience. Front. Neurosci. 7, 83. Kessler, R.C., Berglund, P., Demler, O., Jin, R., Merikangas, K.R., Labuschagne, I., Phan, K.L., Wood, A., Angstadt, M., Chua, P., Walters, E.E., 2005a. Lifetime prevalence and age-of-onset Heinrichs, M., et al., 2010. Oxytocin attenuates amygdala distributions of DSM-IV disorders in the national reactivity to fear in generalized social anxiety disorder. comorbidity survey replication. Arch. Gen. Psychiatry 62 (6), Neuropsychopharmacol.: Off. Publ. Am. Coll. 593–602. Neuropsychopharmacol. 35 (12), 2403–2413. Kessler, R.C., Chiu, W.T., Demler, O., Merikangas, K.R., Walters, E. Labuschagne, I., Phan, K.L., Wood, A., Angstadt, M., Chua, P., E., 2005b. Prevalence, severity, and comorbidity of 12-month Heinrichs, M., et al., 2011. Medial frontal hyperactivity to sad DSM-IV disorders in the national comorbidity survey faces in generalized social anxiety disorder and modulation replication. Arch. Gen. Psychiatry 62 (6), 617–627. by oxytocin. Int. J. Neuropsychopharmacol./Off. Sci. J. Coll. Int. Killgore, W.D., Britton, J.C., Schwab, Z.J., Price, L.M., Weiner, M.R., Neuropsychopharmacol., 1–14 The International Journal of Gold, A.L., et al., 2014. Cortico-limbic responses to masked Neuropsychopharmacology / Volume 15 / Issue 07 / August affective faces across ptsd, panic disorder, and specific 2012, pp 883 896.DOI: 10.1017/S1461145711001489, Published phobia. Depress. Anxiety. 31 (2), 150–159. online: 14 October 2011. Kim, H.S., Sherman, D.K., Sasaki, J.Y., Xu, J., Chu, T.Q., Ryu, C., et Lahoud, N., Maroun, M., 2013. Oxytocinergic manipulations in al., 2010. Culture, distress, and oxytocin receptor corticolimbic circuit differentially affect fear acquisition and polymorphism (OXTR) interact to influence emotional extinction. Psychoneuroendocrinology 38 (10), 2184–2195. support seeking. Proc. Natl. Acad. Sci. USA 107 (36), Lang, P.J., Davis, M., 2006. Emotion, motivation, and the brain: reflex 15717–15721. foundations in animal and human research. Prog. Brain Res. 156, Kim, M.J., Whalen, P.J., 2009. The structural integrity of an 3–29. amygdala-prefrontal pathway predicts trait anxiety. J. Lang, P.J., Davis, M., Ohman, A., 2000. Fear and anxiety: animal Neurosci. 29 (37), 11614–11618. models and human cognitive psychophysiology. J. Affect. Kim, M.J., Loucks, R.A., Palmer, A.L., Brown, A.C., Solomon, K.M., Disord. 61 (3), 137–159. Marchante, A.N., et al., 2011. The structural and Lawson, E.A., Donoho, D.A., Blum, J.I., Meenaghan, E.M., Misra, functional connectivity of the amygdala: from normal M., Herzog, D.B., et al., 2011. Decreased nocturnal oxytocin emotion to pathological anxiety. Behav. Brain Res. 223 (2), levels in anorexia nervosa are associated with low bone 403–410. density and fat mass. J. Clin. Psychiatry 72 (11), Kirsch, P., Esslinger, C., Chen, Q., Mier, D., Lis, S., Siddhanti, S., et 1546–1551. al., 2005. Oxytocin modulates neural circuitry for social Lawson, E.A., Holsen, L.M., Santin, M., DeSanti, R., Meenaghan, E., cognition and fear in humans. J. Neurosci. 25 (49), Eddy, K.T., et al., 2013. Postprandial oxytocin secretion is 11489–11493. associated with severity of anxiety and depressive Kiss, I., Levy-Gigi, E., Keri, S., 2011. CD 38 expression, attachment symptoms in anorexia nervosa. J. Clin. Psychiatry 74 (5), style and habituation of arousal in relation to trust-related e451–457. oxytocin release. Biol. Psychol. 88 (2–3), 223–226. Lawson, E.A., Ackerman, K.E., Estella, N.M., Guereca, G., Pierce, L., Kitano, H., 2004. Biological robustness. Nat. Rev. Genet. 5 (11), Sluss, P.M., et al., 2013. Nocturnal oxytocin secretion is lower 826–837. in amenorrheic athletes than nonathletes and associated with Klumpers, F., van Gerven, J.M., Prinssen, E.P., Niklson, I., Roesch, bone microarchitecture and finite element analysis F., Riedel, W.J., et al., 2010. Method development studies for parameters. Eur. J. Endocrinol. 168 (3), 457–464. repeatedly measuring anxiolytic drug effects in healthy LeDoux, J.E., 1996. The Emotional Brain. Simon and Schuster, New humans. J. Psychopharmacol. 24 (5), 657–666. York. Knobloch, H.S., Charlet, A., Hoffmann, L.C., Eliava, M., Khrulev, S., LeDoux, J., 2003. The emotional brain, fear, and the amygdala. Cetin, A.H., et al., 2012. Evoked axonal oxytocin release in the Cell. Mol. Neurobiol. 23 (4–5), 727–738. central amygdala attenuates fear response. Neuron 73 (3), LeDoux, J.E., Iwata, J., Cicchetti, P., Reis, D.J., 1988. Different 553–566. projections of the central amygdaloid nucleus mediate Koch, S.B.J., Nawijn, L., van Zuiden, M., Frijling, J.L., Veltman, D.J., autonomic and behavioral correlates of conditioned fear. J. Olff, M., 2013. Valence-dependent effects of intranasal Neurosci. 8 (7), 2517–2529. oxytocin on amygdala reactivity in male PTSD patients Leckman, J.F., Goodman, W.K., North, W.G., Chappell, P.B., Price, L. (abstract). Paper Presented at the ISPNEC. H., Pauls, D.L., et al., 1994. Elevated cerebrospinal fluid levels Kojima, S., Stewart, R.A., Demas, G.E., Alberts, J.R., 2012. Maternal of oxytocin in obsessive-compulsive disorder. Comparison contact differentially modulates central and peripheral with Tourette’s syndrome and healthy controls. Arch. Gen. oxytocin in rat pups during a brief regime of mother-pup Psychiatry 51 (10), 782–792. interaction that induces a filial huddling preference. Journal of Lee, R., Ferris, C., Van de Kar, L.D., Coccaro, E.F., 2009. neuroendocrinology 24 (5), 831–840. Cerebrospinal fluid oxytocin, life history of aggression, and 50 brain research 1580 (2014) 22–56

personality disorder. Psychoneuroendocrinology 34 (10), in conditional oxytocin receptor knockout females. Behav. 1567–1573. Neurosci. 124 (5), 677–685. Lee, Y., Davis, M., 1997. Role of the hippocampus, the bed nucleus MacDonald, K., Feifel, D., 2012. Dramatic improvement in sexual of the stria terminalis, and the amygdala in the excitatory function induced by intranasal oxytocin. J. Sex. Med. 9 (5), effect of corticotropin-releasing hormone on the acoustic 1407–1410. startle reflex. J. Neurosci. Off. Soc. Neurosci. 17 (16), 6434–6446. Macdonald, K., Feifel, D., 2013. Helping oxytocin deliver: Leichsenring, F., Leibing, E., Kruse, J., New, A.S., Leweke, F., 2011. considerations in the development of oxytocin-based Borderline personality disorder. Lancet 377 (9759), 74–84. therapeutics for brain disorders. Front. neurosci. 7, 35. Leuner, B., Caponiti, J.M., Gould, E., 2012. Oxytocin stimulates MacDonald, E., Dadds, M.R., Brennan, J.L., Williams, K., Levy, F., adult neurogenesis even under conditions of stress and Cauchi, A.J., 2011. A review of safety, side-effects and elevated . Hippocampus 22 (4), 861–868. subjective reactions to intranasal oxytocin in human Li, Y., Qin, W., Jiang, T., Zhang, Y., Yu, C., 2012. Sex-dependent research. Psychoneuroendocrinology. 36 (8), 1114–1126. correlations between the personality dimension of harm Macdonald, K., Feifel, D., Brune, M., Lamb, K., Wilson, M.P., avoidance and the resting-state functional connectivity of Golshan, S., et al., 2013a. Not disappointed by anxiety: a reply amygdala subregions. PloS One 7 (4), e35925. to Cardoso and Ellenbogen’s commentary “oxytocin and Light, G.A., Swerdlow, N.R., Rissling, A.J., Radant, A., Sugar, C.A., psychotherapy: keeping context and person in mind”. Sprock, J., et al., 2012. Characterization of neurophysiologic and Psychoneuroendocrinology 38 (12), 3173–3175. neurocognitive biomarkers for use in genomic and clinical Macdonald, K., Macdonald, T.M., Brune, M., Lamb, K., Wilson, M. outcome studies of schizophrenia. PloS One 7 (7), e39434. P., Golshan, S., et al., 2013b. Oxytocin and psychotherapy: a Light, K.C., Grewen, K.M., Amico, J.A., 2005. More frequent partner pilot study of its physiological, behavioral and subjective hugs and higher oxytocin levels are linked to lower blood effects in males with depression. Psychoneuroendocrinology pressure and heart rate in premenopausal women. Biol. 38 (12), 2831–2843. Psychol. 69 (1), 5–21. Macdonald, K.S., 2012. Sex, receptors, and attachment: a review Lin, E.J., 2012. Neuropeptides as therapeutic targets in anxiety of individual factors influencing response to oxytocin. Front. disorders. Curr. Pharm. Des. 18 (35), 5709–5727. Neurosci. 6, 194. Linnen, A.M., Ellenbogen, M.A., Cardoso, C., Joober, R., 2012. Mak, P., Broussard, C., Vacy, K., Broadbear, J.H., 2012. Modulation Intranasal oxytocin and salivary cortisol concentrations of anxiety behavior in the elevated plus maze using peptidic during social rejection in university students. Stress 15 (4), oxytocin and vasopressin receptor ligands in the rat. J. 393–402. Psychopharmacol. 26 (4), 532–542. Linnman, C., Moulton, E.A., Barmettler, G., Becerra, L., Borsook, Mamrut, S., Harony, H., Sood, R., Shahar-Gold, H., Gainer, H., Shi, D., 2012. Neuroimaging of the periaqueductal gray: state of the Y.J., et al., 2013. DNA methylation of specific CpG sites in the field. NeuroImage 60 (1), 505–522. promoter region regulates the transcription of the mouse Liu, W., Pappas, G.D., Carter, C.S., 2005. Oxytocin receptors in oxytocin receptor. PloS One 8 (2), e56869. brain cortical regions are reduced in haploinsufficient ( / ) Mantella, R.C., Vollmer, R.R., Li, X., Amico, J.A., 2003. Female þ À reeler mice. Neurol. Res. 27 (4), 339–345. oxytocin-deficient mice display enhanced anxiety-related Lochner, C., Hemmings, S., Seedat, S., Kinnear, C., Schoeman, R., behavior. Endocrinology 144 (6), 2291–2296. Annerbrink, K., et al., 2007. Genetics and personality traits in Marsh, A.A., Yu, H.H., Pine, D.S., Blair, R.J., 2010. Oxytocin patients with social anxiety disorder: a case-control study in improves specific recognition of positive facial expressions. South Africa. Eur. Neuropsychopharmacol. 17 (5), 321–327. Psychopharmacology (Berlin) 209 (3), 225–232. Love, T.M., Enoch, M.A., Hodgkinson, C.A., Pecina, M., Mickey, B., Martin, E.I., Ressler, K.J., Binder, E., Nemeroff, C.B., 2009. The Koeppe, R.A., et al., 2012. Oxytocin gene polymorphisms neurobiology of anxiety disorders: brain imaging, genetics, influence human dopaminergic function in a sex-dependent and psychoneuroendocrinology. Psychiatr. Clin. N. Am. 32 (3), manner. Biol. Psychiatry 72 (3), 198–206. 549–575. Lovero, K.L., Simmons, A.N., Aron, J.L., Paulus, M.P., 2009. Anterior McClure, E.B., Adler, A., Monk, C.S., Cameron, J., Smith, S., Nelson, insular cortex anticipates impending stimulus significance. E.E., et al., 2007. fMRI predictors of treatment outcome in Neuroimage 45 (3), 976–983. pediatric anxiety disorders. Psychopharmacology (Berlin) 191 Lukas, M., Bredewold, R., Neumann, I.D., Veenema, A.H., 2010. (1), 97–105. Maternal separation interferes with developmental changes in McCrory, E., De Brito, S.A., Viding, E., 2012. The link between child brain vasopressin and oxytocin receptor binding in male rats. abuse and psychopathology: a review of neurobiological and Neuropharmacology 58 (1), 78–87. genetic research. J. R. Soc. Med. 105 (4), 151–156. Ludwig, M., Leng, G., 2006. Dendritic peptide release and peptide- McCullough, M.E., Churchland, P.S., Mendez, A.J., 2013. Problems dependent behaviours. Nat. Rev. Neurosci. 7 (2), 126–136. with measuring peripheral oxytocin: can the data on oxytocin Lukas, M., Toth, I., Reber, S.O., Slattery, D.A., Veenema, A.H., and human behavior be trusted?. Neurosci. Biobehav. Rev. 37 Neumann, I.D., 2011. The neuropeptide oxytocin facilitates (8), 1485–1492. pro-social behavior and prevents social avoidance in rats and McGregor, I.S., Bowen, M.T., 2012. Breaking the loop: oxytocin as a mice. Neuropsychopharmacology 36 (11), 2159–2168. potential treatment for drug addiction. Horm. Behav. 61 (3), Lukas, M., Toth, I., Veenema, A.H., Neumann, I.D., 2013. Oxytocin 331–339. mediates rodent social memory within the lateral septum and McKie, S., Del-Ben, C., Elliott, R., Williams, S., del Vai, N., the medial amygdala depending on the relevance of the social Anderson, I., et al., 2005. Neuronal effects of acute citalopram stimulus: male juvenile versus female adult conspecifics. detected by pharmacoMRI. Psychopharmacology (Berlin) 180 Psychoneuroendocrinology 38 (6), 916–926. (4), 680–686. Luther, J.A., Daftary, S.S., Boudaba, C., Gould, G.C., Halmos, K.C., McNair, D.M., Frankenthaler, L.M., Czerlinsky, T., White, T.W., Sasson, Tasker, J.G., 2002. Neurosecretory and non-neurosecretory S., Fisher, S., 1982. Simulated public speaking as a model of parvocellular neurones of the hypothalamic paraventricular clinical anxiety. Psychopharmacology (Berlin) 77 (1), 7–10. nucleus express distinct electrophysiological properties. J. McRae-Clark, A.L., Baker, N.L., Maria, M.M., Brady, K.T., 2013. Neuroendocrinol. 14 (12), 929–932. Effect of oxytocin on craving and stress response in Macbeth, AH., Stepp, J.E., Lee, H.J., Young, W.S., Caldwell, H.K., marijuana-dependent individuals: a pilot study. 2010. Normal maternal behavior, but increased pup mortality, Psychopharmacology. (Berl). 228 (4), 623–631. brain research 1580 (2014) 22–56 51

McNaughton, N., Corr, P.J., 2004. A two-dimensional Neumann, I.D., 2002. Involvement of the brain oxytocin system in neuropsychology of defense: fear/anxiety and defensive stress coping: interactions with the hypothalamo-pituitary- distance. Neurosci. Biobehav. Rev. 28 (3), 285–305. adrenal axis. Prog. Brain Res. 139, 147–162. Meaney, M.J., 2001. Maternal care, gene expression, and the Neumann, I.D., Landgraf, R., 2012. Balance of brain oxytocin and transmission of individual differences in stress reactivity vasopressin: implications for anxiety, depression, and social across generations. Annu. Rev. Neurosci. 24, 1161–1192. behaviors. Trends Neurosci. 35 (11), 649–659. Meaney, M.J., Szyf, M., Seckl, J.R., 2007. Epigenetic mechanisms of Newman, S.W., 1999. The medial extended amygdala in male perinatal programming of hypothalamic-pituitary-adrenal reproductive behavior. A node in the mammalian social function and health. Trends Mol. Med. 13 (7), 269–277. behavior network. Ann. N. Y. Acad. Sci. 877, 242–257. Mehta, M.A., Golembo, N.I., Nosarti, C., Colvert, E., Mota, A., Nitschke, J.B., Sarinopoulos, I., Oathes, D.J., Johnstone, T., Whalen, Williams, S.C., et al., 2009. Amygdala, hippocampal and P.J., Davidson, R.J., et al., 2009. Anticipatory activation in the corpus callosum size following severe early institutional amygdala and anterior cingulate in generalized anxiety deprivation: the English and Romanian adoptees study pilot. J. disorder and prediction of treatment response. Am. J. Child Psychol. Psychiatry 50 (8), 943–951. Psychiatry 166 (3), 302–310. Melzig, C.A., Michalowski, J.M., Holtz, K., Hamm, A.O., 2008. Nolte, T., Guiney, J., Fonagy, P., Mayes, L.C., Luyten, P., 2011. Anticipation of interoceptive threat in highly anxiety sensitive Interpersonal stress regulation and the development of persons. Behav. Res. Ther. 46 (10), 1126–1134. anxiety disorders: an attachment-based developmental Menon, V., 2011. Large-scale brain networks and framework. Front. Behav. Neurosci. 5, 55. psychopathology: a unifying triple network model. Trends Norbury, R., Taylor, M.J., Selvaraj, S., Murphy, S.E., Harmer, C.J., Cogn. Sci. 15 (10), 483–506. Cowen, P.J., 2009. Short-term antidepressant treatment Meyer-Lindenberg, A., Weinberger, D.R., 2006. Intermediate modulates amygdala response to happy faces. phenotypes and genetic mechanisms of psychiatric disorders. Psychopharmacology (Berlin) 206 (2), 197–204. Nat. Rev. Neurosci. 7 (10), 818–827. Norman, G.J., Karelina, K., Morris, J.S., Zhang, N., Cochran, M., Milad, M.R., Rauch, S.L., Pitman, R.K., Quirk, G.J., 2006. Fear Courtney DeVries, A., 2010. Social interaction prevents the extinction in rats: implications for human brain imaging and development of depressive-like behavior post nerve injury in anxiety disorders. Biol. Psychol. 73 (1), 61–71. mice: a potential role for oxytocin. Psychosom. Med. 72 (6), Miller, M., Bales, K.L., Taylor, S.L., Yoon, J., Hostetler, C.M., Carter, 519–526. C.S., et al., 2013. Oxytocin and vasopressin in children and Norrholm, S.D., Ressler, K.J., 2009. Genetics of anxiety and adolescents with autism spectrum disorders: sex differences trauma-related disorders. Neuroscience 164 (1), 272–287. and associations with symptoms. Autism Res.: Off. J. Int. Soc. Okimoto, N., Bosch, O.J., Slattery, D.A., Pflaum, K., Matsushita, H., Autism Res. 6 (2), 91–102. Wei, F.Y., et al., 2012. RGS2 mediates the anxiolytic effect of Milrod, B., Markowitz, J.C., Gerber, A.J., Cyranowski, J., Altemus, oxytocin. Brain Res. 1453, 26–33. M., Shapiro, T., et al., 2013. Childhood separation anxiety and Oler, J.A., Fox, A.S., Shelton, S.E., Rogers, J., Dyer, T.D., Davidson, R. the pathogenesis and treatment of adult anxiety. Am. J. J., et al., 2010. Amygdalar and hippocampal substrates of Psychiatry 171 (1), 34–43. anxious temperament differ in their heritability. Nature 466 Mineka, S., Zinbarg, R., 2006. A contemporary learning theory (7308), 864–868. perspective on the etiology of anxiety disorders: it’s not what Oler, J.A., Birn, R.M., Patriat, R., Fox, A.S., Shelton, S.E., Burghy, C. you thought it was. Am. Psychol. 61 (1), 10–26. A., et al., 2012. Evidence for coordinated functional activity Mobbs, D., Petrovic, P., Marchant, J.L., Hassabis, D., Weiskopf, N., within the extended amygdala of non-human and human Seymour, B., et al., 2007. When fear is near: threat imminence primates. NeuroImage 61 (4), 1059–1066. elicits prefrontal-periaqueductal gray shifts in humans. Olff, M., Langeland, W., Witteveen, A., Denys, D., 2010. A Science 317 (5841), 1079–1083. psychobiological rationale for oxytocin in the treatment of Mobbs, D., Yu, R., Rowe, J.B., Eich, H., FeldmanHall, O., Dalgleish, posttraumatic stress disorder. CNS Spectr. 15 (8), 522–530. T., 2010. Neural activity associated with monitoring the Olsavsky, A.K., Telzer, E.H., Shapiro, M., Humphreys, K.L., oscillating threat value of a tarantula. Proc. Natl. Acad. Sci. Flannery, J., Goff, B., et al., 2013. Indiscriminate amygdala USA 107 (47), 20582–20586. response to mothers and strangers after early maternal Modahl, C., Green, L., Fein, D., Morris, M., Waterhouse, L., deprivation. Biol. Psychiatry 74 (11), 853–860. Feinstein, C., et al., 1998. Plasma oxytocin levels in autistic Ono, Y., Ando, J., Onoda, N., Yoshimura, K., Momose, T., Hirano, children. Biol. Psychiatry 43 (4), 270–277. M., et al., 2002. Dimensions of temperament as vulnerability Moriceau, S., Sullivan, R.M., 2005. Neurobiology of infant factors in depression. Mol. Psychiatry 7 (9), 948–953. attachment. Dev. Psychobiol. 47 (3), 230–242. Opacka-Juffry, J., Mohiyeddini, C., 2012. Experience of stress in Morris, J.S., Scott, S.K., Dolan, R.J., 1999. Saying it with feeling: childhood negatively correlates with plasma oxytocin neural responses to emotional vocalizations. concentration in adult men. Stress 15 (1), 1–10. Neuropsychologia 37 (10), 1155–1163. Ophir, A.G., Gessel, A., Zheng, D.J., Phelps, S.M., 2012. Oxytocin Murgatroyd, C., Spengler, D., 2011. Epigenetic programming of the receptor density is associated with male mating tactics and HPA axis: early life decides. Stress 14 (6), 581–589. social monogamy. Horm. Behav. 61 (3), 445–453. Murphy, S.E., Downham, C., Cowen, P.J., Harmer, C.J., 2008. Direct Otero-Garcia, M., Martin-Sanchez, A., Fortes-Marco, L., Martinez- effects of diazepam on emotional processing in healthy Ricos, J., Agustin-Pavon, C., Lanuza, E., et al., 2013. volunteers. Psychopharmacology (Berlin) 199 (4), 503–513. Extending the socio-sexual brain: arginine-vasopressin Murphy, S.E., Norbury, R., O’Sullivan, U., Cowen, P.J., Harmer, C. immunoreactive circuits in the telencephalon of mice. Brain J., 2009. Effect of a single dose of citalopram on amygdala Struct. Funct. (http://dx.doi.org/10.1007/s00429-013-0553-3, response to emotional faces. Br. J. Psychiatry 194 (6), online). 535–540. Owen, S.F., Tuncdemir, S.N., Bader, P.L., Tirko, N.N., Fishell, G., Myers, K.M., Davis, M., 2007. Mechanisms of fear extinction. Mol. Tsien, R.W., 2013. Oxytocin enhances hippocampal spike Psychiatry 12 (2), 120–150. transmission by modulating fast-spiking interneurons. Nature Nelson, E.E., Panksepp, J., 1998. Brain substrates of infant-mother 500 (7463), 458–462. attachment: contributions of opioids, oxytocin, and Panksepp, J., 1998. Affective Neuroscience. Oxford University norepinephrine. Neurosci. Biobehav. Rev. 22 (3), 437–452. Press, New York. 52 brain research 1580 (2014) 22–56

Panksepp, J., 2006. Emotional endophenotypes in evolutionary Porges, S.W., 1998. Love: an emergent property of the mammalian psychiatry. Prog. Neuropsychopharmacol. Biol. Psychiatry. 30 autonomic nervous system. Psychoneuroendocrinology 23 (8), (5), 774–784. 837–861. Panksepp, J., Northoff, G., 2009. The trans-species core SELF: the Porges, S.W., 2003a. The polyvagal theory: phylogenetic emergence of active cultural and neuro-ecological agents contributions to social behavior. Physiol. Behav. 79 (3), through self-related processing within subcortical-cortical 503–513. midline networks. Conscious. Cogn. 18 (1), 193–215. Porges, S.W., 2003b. Social engagement and attachment: a Panksepp, J., Herman, B., Conner, R., Bishop, P., Scott, J.P., 1978. phylogenetic perspective. Ann. N. Y. Acad. Sci. 1008, 31–47. The biology of social attachments: opiates alleviate separation Porges, S.W., Carter, S.C., 2012. Neurobiology and evolution: distress. Biol. Psychiatry 13 (5), 607–618. mechanisms, mediators, and adaptive consequences of Panksepp, J., Nelson, E., Bekkedal, M., 1997. Brain systems for the caregiving. In: Brown, S.L., Irown, R.M., Penner, L.A. (Eds.), mediation of social separation-distress and social-reward. Moving Beyond Self-Interest: Perspectives From Evolutionary Evolutionary antecedents and neuropeptide intermediaries. Biology, Neuroscience, and the Social Sciences. Oxford Ann. N. Y. Acad. Sci. 807, 78–100. University Press, New York, pp. 53–75. Passingham, R.E., Stephan, K.E., Kotter, R., 2002. The anatomical Pournajafi-Nazarloo, H., Kenkel, W., Mohsenpour, S.R., basis of functional localization in the cortex. Nat. Rev. Sanzenbacher, L., Saadat, H., Partoo, L., et al. (2013). Exposure Neurosci. 3 (8), 606–616. to chronic isolation modulates receptors mRNAs for oxytocin Paulus, M.P., Stein, M.B., 2006. An insular view of anxiety. Biol. and vasopressin in the hypothalamus and heart. Peptides. Psychiatry 60 (4), 383–387. Prehn, K., Kazzer, P., Lischke, A., Heinrichs, M., Herpertz, S.C., Paulus, M.P., Feinstein, J.S., Castillo, G., Simmons, A.N., Stein, M. Domes, G., 2013. Effects of intranasal oxytocin on pupil B., 2005. Dose-dependent decrease of activation in bilateral dilation indicate increased salience of socioaffective stimuli. amygdala and insula by lorazepam during emotion Psychophysiology 50 (6), 528–537. processing. Arch. Gen. Psychiatry 62 (3), 282–288. Quirin, M., Kuhl, J., Dusing, R., 2011. Oxytocin buffers cortisol Pedersen, C.A., Boccia, M.L., 2002. Oxytocin links mothering responses to stress in individuals with impaired emotion received, mothering bestowed and adult stress responses. regulation abilities. Psychoneuroendocrinology 36 (6), Stress 5 (4), 259–267. 898–904. Pedersen, C.A., Prange Jr., A.J., 1979. Induction of maternal behavior Radke, S., Roelofs, K., de Bruijn, E.R., 2013. Acting on anger: in virgin rats after intracerebroventricular administration of social anxiety modulates approach-avoidance tendencies oxytocin. Proc. Natl. Acad. Sci. USA 76 (12), 6661–6665. after oxytocin administration. Psychol. Sci. 24 (8), Pedersen, C.A., Vadlamudi, S.V., Boccia, M.L., Amico, J.A., 2006. 1573–1578. Maternal behavior deficits in nulliparous oxytocin knockout Raineki, C., Moriceau, S., Sullivan, R.M., 2010. Developing a mice. Genes Brain Behav. 5 (3), 274–281. neurobehavioral animal model of infant attachment to an Pedersen, C.A., Smedley, K.L., Leserman, J., Jarskog, L.F., Rau, S.W., abusive caregiver. Biol. Psychiatry 67 (12), 1137–1145. Kampov-Polevoi, A., et al., 2013. Intranasal oxytocin blocks Ravinder, S., Burghardt, N.S., Brodsky, R., Bauer, E.P., Chattarji, S., alcohol withdrawal in human subjects. Alcohol. Clin. Exp. Res. 2013. A role for the extended amygdala in the fear-enhancing 37 (3), 484–489. effects of acute selective serotonin reuptake inhibitor Pego, J.M., Morgado, P., Pinto, L.G., Cerqueira, J.J., Almeida, O.F., treatment. Transl. Psychiatry 3, e209. Sousa, N., 2008. Dissociation of the morphological correlates Ravindran, L.N., Stein, M.B., 2010. The pharmacologic treatment of stress-induced anxiety and fear. Eur. J. Neurosci. 27 (6), of anxiety disorders: a review of progress. J. Clin. Psychiatry 71 1503–1516. (7), 839–854. Petrovic, P., Kalisch, R., Singer, T., Dolan, R.J., 2008. Oxytocin Risbrough, V., 2010. Behavioral correlates of anxiety. Curr. Top. attenuates affective evaluations of conditioned faces and Behav. Neurosci. 2, 205–228. amygdala activity. J. Neurosci. 28 (26), 6607–6615. Ribary, U., 2005. Dynamics of thalamo-cortical network oscillations Peters, J., Kalivas, P.W., Quirk, G.J., 2009. Extinction circuits for fear and human perception. Prog. Brain Res. 150, 127–142. and addiction overlap in prefrontal cortex. Learn. & Mem. 16 Richter-Levin, G., Maroun, M., 2010. Stress and amygdala (5), 279–288. suppression of metaplasticity in the medial prefrontal cortex. Phan, K.L., Coccaro, E.F., Angstadt, M., Kreger, K.J., Mayberg, H.S., Cereb. Cortex 20 (10), 2433–2441. Liberzon, I., et al., 2013. Corticolimbic brain reactivity to Riem, M.M., Bakermans-Kranenburg, M.J., Pieper, S., Tops, M., social signals of threat before and after sertraline treatment Boksem, M.A., Vermeiren, R.R., et al., 2011. Oxytocin in generalized social phobia. Biol. Psychiatry 73 (4), modulates amygdala, insula, and inferior frontal gyrus 329–336. responses to infant crying: a randomized controlled trial. Biol. Pierrehumbert, B., Torrisi, R., Ansermet, F., Borghini, A., Halfon, O., Psychiatry 70 (3), 291–297. 2012. Adult attachment representations predict cortisol and Riem, M.M., van, I.M.H., Tops, M., Boksem, M.A., Rombouts, S.A., oxytocin responses to stress. Attach. Hum. Dev. 14 (5), 453–476. Bakermans-Kranenburg, M.J., 2012. No laughing matter: Pincus, D., Kose, S., Arana, A., Johnson, K., Morgan, P.S., intranasal oxytocin administration changes functional brain Borckardt, J., et al., 2010. Inverse effects of oxytocin on connectivity during exposure to infant laughter. attributing mental activity to others in depressed and healthy Neuropsychopharmacol.: Off. Publ. Am. Coll. subjects: a double-blind placebo controlled FMRI study. Front. Neuropsychopharmacol. 37 (5), 1257–1266. Psychiatry/Front. Res. Found. 1, 134. Riem, M.M., van Ijzendoorn, M.H., Tops, M., Boksem, M.A., Pitman, R.K., Orr, S.P., Lasko, N.B., 1993. Effects of intranasal Rombouts, S.A., Bakermans-Kranenburg, M.J., 2013. Oxytocin vasopressin and oxytocin on physiologic responding during effects on complex brain networks are moderated by personal combat imagery in Vietnam veterans with experiences of maternal love withdrawal. Eur. posttraumatic stress disorder. Psychiatry Res. 48 (2), 107–117. Neuropsychopharmacol.: J. Eur. Coll. Neuropsychopharmacol. Poirier, G.L., Cordero, M.I., Sandi, C., 2013. Female vulnerability to 23 (10), 1288–1295. the development of depression-like behavior in a rat model of Rilling, J.K., 2013. The neural and hormonal bases of human intimate partner violence is related to anxious temperament, parental care. Neuropsychologia 51 (4), 731–747. coping responses, and amygdala vasopressin receptor 1a Ring, R.H., Schechter, L.E., Leonard, S.K., Dwyer, J.M., Platt, B.J., expression. Front. Behav. Neurosci. 7, 35. Graf, R., et al., 2010. Receptor and behavioral pharmacology of brain research 1580 (2014) 22–56 53

WAY-267464, a non-peptide oxytocin receptor agonist. underpins empathic communication difficulties at the first Neuropharmacology 58 (1), 69–77. stages of romantic love. Soc. Cogn. Affect. Neurosci.. Ring Atzil, S., Hendler, T., Feldman, R., 2011. Specifying the Schorscher-Petcu, A., Sotocinal, S., Ciura, S., Dupre, A., Ritchie, J., neurobiological basis of human attachment: brain, hormones, Sorge, R.E., et al., 2010. Oxytocin-induced analgesia and and behavior in synchronous and intrusive mothers. scratching are mediated by the vasopressin-1A receptor in the Neuropsychopharmacology. 36 (13), 2603–2615. mouse. J. Neurosci. 30 (24), 8274–8284. Robinson, M.D., 1998. Running from William James’ Bear: a Schwartz, C.E., Kunwar, P.S., Greve, D.N., Kagan, J., Snidman, N.C., review of preattentive mechanism and their contribution to Bloch, R.B., 2012. A phenotype of early infancy predicts reactivity emotional experience. Cogn. Emot. 12 (5), 667–696. of the amygdala in male adults. Mol. Psychiatry 17 (10), 1042–1050. Rodrigues, S.M., Saslow, L.R., Garcia, N., John, O.P., Keltner, D., Scantamburlo, G., Ansseau, M., Geenen, V., Legros, J.J., 2011. 2009. Oxytocin receptor genetic variation relates to empathy Intranasal oxytocin as an adjunct to escitalopram in major and stress reactivity in humans. Proc. Natl. Acad. Sci. USA 106 depression. J. Neuropsychiatry. Clin. Neurosci. 23 (2), E5. (50), 21437–21441. Seeley, W.W., Menon, V., Schatzberg, A.F., Keller, J., Glover, G.H., Rosen, G.J., de Vries, G.J., Goldman, S.L., Goldman, B.D., Forger, N. Kenna, H., et al., 2007. Dissociable intrinsic connectivity G., 2008. Distribution of oxytocin in the brain of a eusocial networks for salience processing and executive control. J. rodent. Neuroscience 155 (3), 809–817. Neurosci. 27 (9), 2349–2356. Rosenbaum, J.F., Biederman, J., Bolduc-Murphy, E.A., Faraone, S.V., Sehlmeyer, C., Dannlowski, U., Schoning, S., Kugel, H., Pyka, M., Chaloff, J., Hirshfeld, D.R., et al., 1993. Behavioral inhibition in Pfleiderer, B., et al., 2011. Neural correlates of trait anxiety in childhood: a risk factor for anxiety disorders. Harv. Rev. fear extinction. Psychol. Med. 41 (4), 789–798. Psychiatry 1 (1), 2–16. Seltzer, L.J., Ziegler, T.E., Pollak, S.D., 2010. Social vocalizations Ross, H.E., Freeman, S.M., Spiegel, L.L., Ren, X., Terwilliger, E.F., can release oxytocin in humans. Proc. Biol. Sci./R. Soc. 277 Young, L.J., 2009. Variation in oxytocin receptor density in the (1694), 2661–2666. nucleus accumbens has differential effects on affiliative Selye, H., 1936. A syndrome produced by diverse nocuous agents. behaviors in monogamous and polygamous voles. J. Neurosci. Nature 138, 32. 29 (5), 1312–1318. Shahrestani, S., Kemp, A.H., Guastella, A.J., 2013. The impact of a Rotzinger, S., Lovejoy, D.A., Tan, L.A., 2010. Behavioral effects of single administration of intranasal oxytocin on the neuropeptides in rodent models of depression and anxiety. recognition of basic emotions in humans: a meta-analysis. Peptides 31 (4), 736–756. Neuropsychopharmacol.: Off. Publ. Am. Coll. Rubin, L.H., Carter, C.S., Drogos, L., Pournajafi-Nazarloo, H., Neuropsychopharmacol. 38 (10), 1929–1936. Sweeney, J.A., Maki, P.M., 2010. Peripheral oxytocin is Shin, L.M., Liberzon, I., 2010. The neurocircuitry of fear, stress, associated with reduced symptom severity in schizophrenia. and anxiety disorders. Neuropsychopharmacol.: Off. Publ. Schizophr. Res. 124 (1–3), 13–21. Am. Coll. Neuropsychopharmacol. 35 (1), 169–191. Rubin, L.H., Carter, C.S., Bishop, J.R., Pournajafi-Nazarloo, H., Shultz, S., Dunbar, R.I., 2007. The evolution of the social brain: Harris, M.S., Hill, S.K., et al., 2013. Peripheral vasopressin but anthropoid primates contrast with other vertebrates. Proc. not oxytocin relates to severity of acute psychosis in women Biol. Sci. 274 (1624), 2429–2436. with acutely-ill untreated first-episode psychosis. Schizophr. Simeon, D., Bartz, J., Hamilton, H., Crystal, S., Braun, A., Ketay, S., Res. 146 (1–3), 138–143. et al., 2011. Oxytocin administration attenuates stress Rupp, H.A., James, T.W., Ketterson, E.D., Sengelaub, D.R., Ditzen, reactivity in borderline personality disorder: a pilot study. B., Heiman, J.R., 2014. Amygdala response to negative images Psychoneuroendocrinology 36 (9), 1418–1421. in postpartum vs nulliparous women and intranasal oxytocin. Skrundz, M., Bolten, M., Nast, I., Hellhammer, D.H., Soc. Cogn. Affect. Neurosci. 9 (1), 48–54. Meinlschmidt, G., 2011. Plasma oxytocin concentration during Ruys, K.I., Stapel, D.A., 2008. The secret life of emotions. Psychol. pregnancy is associated with development of postpartum Sci. 19 (4), 385–391. depression. Neuropsychopharmacol.: Off. Publ. Am. Coll. Salzberg, H.C., Lonstein, J.S., Stern, J.M., 2002. GABA(A) receptor Neuropsychopharmacol. 36 (9), 1886–1893. regulation of kyphotic nursing and female sexual behavior in Skopek, P., Hynie, S., Chottova-Dvorakova, M., Sida, P., Slavikova, J., the caudal ventrolateral periaqueductal gray of postpartum Mistrova, E., et al., 2012. Effects of acute stressors on the rats. Neuroscience 114 (3), 675–687. expression of oxytocin receptor mRNA in hearts of rats with Sasayama, D., Hattori, K., Teraishi, T., Hori, H., Ota, M., Yoshida, different activity of HPA axis. Neuro Endocrinol. Lett., 33; 123–132. S., et al., 2012. Negative correlation between cerebrospinal Slattery, D.A., Neumann, I.D., 2008. No stress please! Mechanisms fluid oxytocin levels and negative symptoms of male of stress hyporesponsiveness of the maternal brain. J. Physiol. patients with schizophrenia. Schizophr. Res. 139 (1–3), 586 (2), 377–385. 201–206. Slattery, D.A., Neumann, I.D., 2010. Chronic icv oxytocin Sawchenko, P.E., Swanson, L.W., 1983. The organization and attenuates the pathological high anxiety state of selectively biochemical specificity of afferent projections to the bred Wistar rats. Neuropharmacology 58 (1), 56–61. paraventricular and supraoptic nuclei. Prog. Brain Res. 60, Smeltzer, M.D., Curtis, J.T., Aragona, B.J., Wang, Z., 2006. 19–29. Dopamine, oxytocin, and vasopressin receptor binding in the Saxena, S., Rauch, S.L., 2000. Functional neuroimaging and the medial prefrontal cortex of monogamous and promiscuous neuroanatomy of obsessive-compulsive disorder. Psychiatr. voles. Neurosci. Lett. 394 (2), 146–151. Clin. N. Am. 23 (3), 563–586. Smith, A.S., Wang, Z., 2013. Hypothalamic oxytocin mediates Schienle, A., Schafer, A., Stark, R., Vaitl, D., 2009. Long-term social buffering of the stress response. Biol. Psychiatry http:// effects of cognitive behavior therapy on brain activation in dx.doi.org/10.1016/j.biopsych.2013.1009.1017. spider phobia. Psychiatry Res. 172 (2), 99–102. Sofroniew, M.V., 1983. Morphology of vasopressin and oxytocin Schneiderman, I., Zagoory-Sharon, O., Leckman, J.F., Feldman, R., neurones and their central and vascular projections. Prog. 2012. Oxytocin during the initial stages of romantic Brain Res. 60, 101–114. attachment: relations to couples’ interactive reciprocity. Soliman, F., Glatt, C.E., Bath, K.G., Levita, L., Jones, R.M., Pattwell, Psychoneuroendocrinology 37 (8), 1277–1285. S.S., et al., 2010. A genetic variant BDNF polymorphism alters Schneiderman, I., Kanat-Maymon, Y., Ebstein, R.P., Feldman, R., extinction learning in both mouse and human. Science 327 2013. Cumulative risk on the oxytocin receptor gene (OXTR) (5967), 863–866. 54 brain research 1580 (2014) 22–56

Somerville, L.H., Whalen, P.J., Kelley, W.M., 2010. Human bed dysfunction in anxiety and anxiety disorders. Trends nucleus of the stria terminalis indexes hypervigilant threat Neurosci. 35 (9), 527–535. monitoring. Biol. Psychiatry 68 (5), 416–424. Szeto, A., McCabe, P.M., Nation, D.A., Tabak, B.A., Rossetti, M.A., Spampinato, M.V., Wood, J.N., De Simone, V., Grafman, J., 2009. McCullough, M.E., et al., 2011. Evaluation of enzyme Neural correlates of anxiety in healthy volunteers: a voxel- immunoassay and radioimmunoassay methods for the based morphometry study. J. Neuropsychiatry Clin. Neurosci. measurement of plasma oxytocin. Psychosom. Med. 73 (5), 21 (2), 199–205. 393–400. Spielberger, C.D., 1983. State-Trait Anxiety Inventory for Adults: Tabak, B.A., McCullough, M.E., Szeto, A., Mendez, A.J., McCabe, P.M., Mind Garden. 2011. Oxytocin indexes relational distress following interpersonal Sporns, O., Honey, C.J., Kotter, R., 2007. Identification and harms in women. Psychoneuroendocrinology 36 (1), classification of hubs in brain networks. PLoS One 2 (10), e1049. 115–122. Sripada, C.S., Phan, K.L., Labuschagne, I., Welsh, R., Nathan, P.J., Tasker, R.R., 1982. Identification of pain processing systems by Wood, A.G., 2012. Oxytocin enhances resting-state electrical stimulation of the brain. Hum. Neurobiol. 1 (4), connectivity between amygdala and medial frontal cortex. Int. 261–272. J. Neuropsychopharmacol./Off. Sci. J. Coll. Int. Taylor, S.E., Klein, L.C., Lewis, B.P., Gruenewald, T.L., Gurung, R.A., Neuropsychopharmacol., 1–6 http://dx.doi.org/10.1017/ Updegraff, J.A., 2000. Biobehavioral responses to stress in S1461145712000533. females: tend-and-befriend, not fight-or-flight. Psychol. Rev. Stankova, T., Eichhammer, P., Langguth, B., Sand, P.G., 2012. 107, 411–429. Sexually dimorphic effects of oxytocin receptor gene (OXTR) Taylor, S.E., Gonzaga, G.C., Klein, L.C., Hu, P., Greendale, G.A., variants on harm avoidance. Biol. Sex Differ. 3 (1), 17. Seeman, T.E., 2006. Relation of oxytocin to psychological Starcevic, V., Uhlenhuth, E.H., 1996. Personality dimensions in stress responses and hypothalamic-pituitary-adrenocortical panic disorder before and after effective treatment. Anxiety 2 axis activity in older women. Psychosom. Med. 68 (2), 238–245. (2), 95–98. Taylor, S.E., Saphire-Bernstein, S., Seeman, T.E., 2010. Are plasma Stein, M.B., Simmons, A.N., Feinstein, J.S., Paulus, M.P., 2007. oxytocin in women and plasma vasopressin in men biomarkers Increased amygdala and insula activation during emotion of distressed pair-bond relationships?. Psychol. Sci. 21 (1), 3–7. processing in anxiety-prone subjects. Am. J. Psychiatry 164 (2), Terasawa, Y., Shibata, M., Moriguchi, Y., Umeda, S., 2013. Anterior 318–327. insular cortex mediates bodily sensibility and social anxiety. Stein, M.B., Schork, N.J., Gelernter, J., 2008. Gene-by-environment Soc. Cogn. Affect. Neurosci. 8 (3), 259–266. (serotonin transporter and childhood maltreatment) Thompson, R.J., Parker, K.J., Hallmayer, J.F., Waugh, C.E., Gotlib, I. interaction for anxiety sensitivity, an intermediate phenotype H., 2011. Oxytocin receptor gene polymorphism (rs2254298) for anxiety disorders. Neuropsychopharmacol.: Off. Publ. Am. interacts with familial risk for psychopathology to predict Coll. Neuropsychopharmacol. 33 (2), 312–319. symptoms of depression and anxiety in adolescent girls. Steuwe, C., Daniels, J.K., Frewen, P.A., Densmore, M., Pannasch, S., Psychoneuroendocrinology 36 (1), 144–147. Beblo, T., et al., 2014. Effect of direct eye contact in PTSD Tieri, P., Grignolio, A., Zaikin, A., Mishto, M., Remondini, D., related to interpersonal trauma: an fMRI study of activation of Castellani, G.C., et al., 2010. Network, degeneracy and bow an innate alarm system. Soc. Cogn. Affect. Neurosci. 9 (1), tie integrating paradigms and architectures to grasp 88–97. the complexity of the immune system. Theor. Biol. Med. Stewart, S.A., 2005. The effects of benzodiazepines on cognition. J. Model. 7, 32. Clin. Psychiatry 66 (Suppl. 2), S9–S13. Todorov, A., 2008. Evaluating faces on trustworthiness: an extension Stolzenberg, D.S., Stevens, J.S., Rissman, E.F., 2012. Experience- of systems for recognition of emotions signaling approach/ facilitated improvements in pup retrieval; evidence for an avoidance behaviors. Ann. N. Y. Acad. Sci. 1124, 208–224. epigenetic effect. Horm. Behav. 62 (2), 128–135. Tol, W.A., Barbui, C., van Ommeren, M., 2013. Management of acute Stoop, R., 2012. Neuromodulation by oxytocin and vasopressin. stress, PTSD, and bereavement: WHO recommendations. JAMA: J. Neuron 76 (1), 142–159. Am. Medical Assoc. 310 (5), 477–478. Strathearn, L., Fonagy, P., Amico, J., Montague, P.R., 2009. Adult Tomasello, M., Carpenter, M., Call, J., Behne, T., Moll, H., 2005. attachment predicts maternal brain and oxytocin response to Understanding and sharing intentions: the origins of cultural infant cues. Neuropsychopharmacology 34 (13), cognition. Behav. Brain Sci. 28 (5), 675–691 (discussion 691–735). 2655–2666. Tomasello, M., Hare, B., Lehmann, H., Call, J., 2007. Reliance on Straube, T., Mentzel, H.J., Miltner, W.H., 2007. Waiting for spiders: head versus eyes in the gaze following of great apes and brain activation during anticipatory anxiety in spider phobics. human infants: the cooperative eye hypothesis. J. Hum. Evol. NeuroImage 37 (4), 1427–1436. 52 (3), 314–320. Striepens, N., Scheele, D., Kendrick, K.M., Becker, B., Schafer, L., Tops, M., Buisman-Pijlman, F.T., Boksem, M.A., Wijers, A.A., Korf, Schwalba, K., et al., 2012. Oxytocin facilitates protective J., 2012. Cortisol-induced increases of plasma oxytocin levels responses to aversive social stimuli in males. Proc. Natl. Acad. predict decreased immediate free recall of unpleasant words. Sci. USA 109 (44), 18144–18149. Front. Psychiatry/Front. Res. Found. 3, 43. Stuebe, A.M., Grewen, K., Meltzer-Brody, S., 2013. Association Tost, H., Kolachana, B., Hakimi, S., Lemaitre, H., Verchinski, B.A., between maternal mood and oxytocin response to Mattay, V.S., et al., 2010. A common allele in the oxytocin breastfeeding. J. Women’s Health 22 (4), 352–361. receptor gene (OXTR) impacts prosocial temperament and Swanson, L.W., 2000. Cerebral hemisphere regulation of human hypothalamic-limbic structure and function. Proc. motivated behavior. Brain Res. 886 (1–2), 113–164. Natl. Acad. Sci. USA 107 (31), 13936–13941. Swanson, L.W., Sawchenko, P.E., 1980. Paraventricular nucleus: a Tost, H., Kolachana, B., Verchinski, B.A., Bilek, E., Goldman, A.L., site for the integration of neuroendocrine and autonomic Mattay, V.S., et al., 2011. Neurogenetic effects of OXTR mechanisms. Neuroendocrinology 31 (6), 410–417. rs2254298 in the extended limbic system of healthy Caucasian Swanson, L.W., Sawchenko, P.E., 1983. Hypothalamic integration: adults. Biol. Psychiatry 70 (9), e37–e39 (author reply e41-32). organization of the paraventricular and supraoptic nuclei. Toth, I., Neumann, I.D., Slattery, D.A., 2012. Central Annu. Rev. Neurosci. 6, 269–324. administration of oxytocin receptor ligands affects cued fear Sylvester, C.M., Corbetta, M., Raichle, M.E., Rodebaugh, T.L., extinction in rats and mice in a timepoint-dependent manner. Schlaggar, B.L., Sheline, Y.I., et al., 2012. Functional network Psychopharmacology (Berlin) 223 (2), 149–158. brain research 1580 (2014) 22–56 55

Tottenham, N., Hare, T.A., Quinn, B.T., McCarry, T.W., Nurse, M., year course of depressive and anxiety disorders. Gilhooly, T., et al., 2010. Prolonged institutional rearing is Psychoneuroendocrinology 38 (9), 1494–1502. associated with atypically large amygdala volume and Wachleski, C., Salum, G.A., Blaya, C., Kipper, L., Paludo, A., difficulties in emotion regulation. Dev. Sci. 13 (1), 46–61. Salgado, A.P., et al., 2008. Harm avoidance and self- Tracey, I., Ploghaus, A., Gati, J.S., Clare, S., Smith, S., Menon, R.S., directedness as essential features of panic disorder patients. et al., 2002. Imaging attentional modulation of pain in the Compr. Psychiatry 49 (5), 476–481. periaqueductal gray in humans. J. Neurosci. 22 (7), 2748–2752. Wager, T.D., van Ast, V.A., Hughes, B.L., Davidson, M.L., Lindquist, M. Turner, R.A., Altemus, M., Enos, T., Cooper, B., McGuinness, T., A., Ochsner, K.N., 2009. Brain mediators of cardiovascular 1999. Preliminary research on plasma oxytocin in normal responses to social threat, Part II: prefrontal-subcortical pathways cycling women: investigating emotion and interpersonal and relationship with anxiety. NeuroImage 47 (3), 836–851. distress. Psychiatry 62 (2), 97–113. Waldman, I.D., Nigg, J.T., Gizer, I.R., Park, L., Rappley, M.D., Tyrka, A.R., Wier, L.M., Price, L.H., Rikhye, K., Ross, N.S., Friderici, K., 2006. The adrenergic receptor alpha-2A gene Anderson, G.M., et al., 2008. Cortisol and ACTH responses to (ADRA2A) and neuropsychological executive functions as the Dex/CRH test: influence of temperament. Horm. Behav. 53 putative endophenotypes for childhood ADHD. Cogn. Affect. (4), 518–525. Behav. Neurosci. 6 (1), 18–30. Unternaehrer, E., Luers, P., Mill, J., Dempster, E., Meyer, A.H., Walker, D.L., Davis, M., 2008. Role of the extended amygdala in Staehli, S., et al., 2012. Dynamic changes in DNA methylation short-duration versus sustained fear: a tribute to Dr. Lennart of stress-associated genes (OXTR, BDNF) after acute Heimer. Brain Struct. Funct. 213 (1–2), 29–42. psychosocial stress. Transl. Psychiatry 2, e150. Walum, H., Westberg, L., Henningsson, S., Neiderhiser, J.M., Reiss, Valassi, E., Scacchi, M., Cavagnini, F., 2008. Neuroendocrine control D., Igl, W., et al., 2008. Genetic variation in the vasopressin of food intake. Nutr. Metab. Cardiovasc. Dis. 18 (2), 158–168. receptor 1a gene (AVPR1A) associates with pair-bonding Van IJzendoorn, M.H., Bakermans-Kranenburg, M.J., 2012. A sniff behavior in humans. Proc Natl Acad Sci USA 105 (37), of trust: meta-analysis of the effects of intranasal oxytocin 14153–14156. administration on face recognition, trust to in-group, and Wang, H., Duclot, F., Liu, Y., Wang, Z., Kabbaj, M., 2013. Histone trust to out-group. Psychoneuroendocrinology 37 (3), 438–443. deacetylase inhibitors facilitate partner preference formation Van Schuerbeek, P., Baeken, C., De Raedt, R., De Mey, J., Luypaert, in female prairie voles. Nat. Neurosci. 16 (7), 919–924. R., 2011. Individual differences in local gray and white matter Wang, J., Qin, W., Liu, B., Zhou, Y., Wang, D., Zhang, Y., et al., 2013. volumes reflect differences in temperament and character: a Neural mechanisms of oxytocin receptor gene mediating voxel-based morphometry study in healthy young females. anxiety-related temperament. Brain Struct. Funct. 26 (May), Brain Res. 1371, 32–42. 2013, http://dx.doi.org/10.1007/s00429-013-0584-9 (Published Veenema, A.H. (2011). Toward understanding how early-life social online). experiences alter oxytocin- and vasopressin-regulated social Wang, J., Qin, W., Liu, B., Wang, D., Zhang, Y., Jiang, T., et al., behaviors. Hormones Behav.. 2013a. Variant in OXTR gene and functional connectivity of Veenema, A.H., 2012. Toward understanding how early-life social the hypothalamus in normal subjects. NeuroImage 81, experiences alter oxytocin- and vasopressin-regulated social 199–204. behaviors. Horm. Behav. 61 (3), 304–312. Wang, J., Qin, W., Liu, B., Zhou, Y., Wang, D., Zhang, Y., et al., Veening, J.G., de Jong, T., Barendregt, H.P., 2010. Oxytocin- 2013b. Neural mechanisms of oxytocin receptor gene messages via the cerebrospinal fluid: behavioral effects; a mediating anxiety-related temperament. Brain Struct. Funct. review. Physiol. Behav. 101 (2), 193–210. Published online 26 May 2013.http://dx.doi.org/10.1007/ Veer, I.M., Oei, N.Y., Spinhoven, P., van Buchem, M.A., Elzinga, B.M., s00429-013-0584-9. Rombouts, S.A., 2012. Endogenous cortisol is associated with Wang, Y., Xu, L., Pan, Y., Wang, Z., Zhang, Z., 2013. Species functional connectivity between the amygdala and medial differences in the immunoreactive expression of oxytocin, prefrontal cortex. Psychoneuroendocrinology 37 (7), 1039–1047. vasopressin, tyrosine hydroxylase and estrogen receptor Veinante, P., Freund-Mercier, M.J., 1997. Distribution of oxytocin- alpha in the brain of Mongolian gerbils (Meriones unguiculatus) and vasopressin-binding sites in the rat extended amygdala: a and Chinese striped hamsters (Cricetulus barabensis). PloS One histoautoradiographic study. J. Comp. Neurol. 383 (3), 305–325. 8 (6), e65807. Ventura-Silva, A.P., Pego, J.M., Sousa, J.C., Marques, A.R., Warren, S.L., Huston, L., Egeland, B., Sroufe, L.A., 1997. Child and Rodrigues, A.J., Marques, F., et al., 2012. Stress shifts the adolescent anxiety disorders and early attachment. J. Am. response of the bed nucleus of the stria terminalis to an Acad. Child Adolesc. Psychiatry 36 (5), 637–644. anxiogenic mode. Eur. J. Neurosci. 36 (10), 3396–3406. Watson, D., 2005. Rethinking the mood and anxiety disorders: a Viviani, D., Stoop, R., 2008. Opposite effects of oxytocin and quantitative hierarchical model for DSM-V. J. Abnorm. vasopressin on the emotional expression of the fear response. Psychol. 114 (4), 522–536. Prog. Brain Res. 170, 207–218. Weisman, O., Zagoory-Sharon, O.,Schneiderman,I.,Gordon,I., Viviani, D., Terrettaz, T., Magara, F., Stoop, R., 2010. Oxytocin Feldman, R., 2013a. Plasma oxytocin distributions in a large enhances the inhibitory effects of diazepam in the rat central cohort of women and men and their gender-specific associations medial amygdala. Neuropharmacology 58 (1), 62–68. with anxiety. Psychoneuroendocrinology 38 (5), 694–701. Viviani, D., Charlet, A., van den Burg, E., Robinet, C., Hurni, N., Weisman, O., Zagoory-Sharon, O., & Feldman, R. (2013). Oxytocin Abatis, M., et al., 2011. Oxytocin selectively gates fear administration alters HPA reactivity in the context of parent– responses through distinct outputs from the central infant interaction. Eur. Neuropsychopharmacol.: J. Eur. Coll. amygdala. Science 333 (6038), 104–107. Neuropsychopharmacol. Volkow, N.D., Wang, G.J., Tomasi, D., Baler, R.D., 2013. The addictive Westlye, L.T., Bjornebekk, A., Grydeland, H., Fjell, A.M., Walhovd, dimensionality of obesity. Biol. Psychiatry 73 (9), 811–818. K.B., 2011. Linking an anxiety-related personality trait to brain Vreeburg, S.A., Zitman, F.G., van Pelt, J., Derijk, R.H., Verhagen, J. white matter microstructure: diffusion tensor imaging and C., van Dyck, R., et al., 2010. Salivary cortisol levels in persons harm avoidance. Arch. Gen. Psychiatry 68 (4), 369–377. with and without different anxiety disorders. Psychosom. Whalen, P.J., Johnstone, T., Somerville, L.H., Nitschke, J.B., Polis, S., Med. 72 (4), 340–347. Alexander, A.L., et al., 2008a. A functional magnetic resonance Vreeburg, S.A., Hoogendijk, W.J., DeRijk, R.H., van Dyck, R., Smit, J. imaging predictor of treatment response to venlafaxine in H., Zitman, F.G., et al., 2013. Salivary cortisol levels and the 2- generalized anxiety disorder. Biol. Psychiatry 63 (9), 858–863. 56 brain research 1580 (2014) 22–56

Whalen, P.J., Johnstone, T., Somerville, L.H., Nitschke, J.B., Polis, S., Zak, P.J., Kurzban, R., Matzner, W.T., 2004. The neurobiology of Alexander, A.L., et al., 2008b. A functional magnetic resonance trust. Ann. N. Y. Acad. Sci. 1032, 224–227. imaging predictor of treatment response to venlafaxine in Zangara, A., Blair, R.J., Curran, H.V., 2002. A comparison of the generalized anxiety disorder. Biol. Psychiatry 63 (9), 858–863. effects of a beta-adrenergic blocker and a benzodiazepine White, M.G., Bogdan, R., Fisher, P.M., Munoz, K.E., Williamson, D. upon the recognition of human facial expressions. E., Hariri, A.R., 2012. FKBP5 and emotional neglect interact to Psychopharmacology (Berlin) 163 (1), 36–41. predict individual differences in amygdala reactivity. Genes Zhang, G., Zhang, Y., Fast, D.M., Lin, Z., Steenwyk, R., 2011. Ultra Brain. Behav. 11 (7), 869–878. sensitive quantitation of endogenous oxytocin in rat and Wu, Z., Xu, Y., Zhu, Y., Sutton, A.K., Zhao, R., Lowell, B.B., et al., human plasma using a two-dimensional liquid 2012. An obligate role of oxytocin neurons in diet induced chromatography-tandem mass spectrometry assay. Anal. energy expenditure. PLoS One 7 (9), e45167. Biochem. 416 (1), 45–52. Yamashita, K., Kitano, T., 2013. Molecular evolution of the Zink, C.F., Meyer-Lindenberg, A., 2012. Human neuroimaging of oxytocin-oxytocin receptor system in eutherians. Mol. oxytocin and vasopressin in social cognition. Horm. Behav. 61 Phylogenet. Evol. 67 (2), 520–528. (3), 400–409. Yamasue, H., Abe, O., Suga, M., Yamada, H., Inoue, H., Tochigi, M., Zink, C.F., Stein, J.L., Kempf, L., Hakimi, S., Meyer-Lindenberg, A., et al., 2008. Gender-common and -specific neuroanatomical 2010. Vasopressin modulates medial prefrontal cortex- basis of human anxiety-related personality traits. Cereb. amygdala circuitry during emotion processing in humans. J. Cortex 18 (1), 46–52. Yang, J., Liang, J.Y., Li, P., Pan, Y.J., Qiu, P.Y., Zhang, J., et al., 2011. Neurosci.: Off. J. Soc. Neurosci. 30 (20), 7017–7022. Oxytocin in the periaqueductal gray participates in pain Zuardi, A.W., Cosme, R.A., Graeff, F.G., Guimaraes, F.S., 1993. modulation in the rat by influencing endogenous opiate Effects of ipsapirone and cannabidiol on human experimental peptides. Peptides 32 (6), 1255–1261. anxiety. J. Psychopharmacol. 7 (Suppl. 1), S82–S88. Yang, T.T., Simmons, A.N., Matthews, S.C., Tapert, S.F., Frank, G. Zukerman, M., Cloninger, C.R., 1996. Relationships between K., Bischoff-Grethe, A., et al., 2009. Adolescent subgenual Cloninger’s, Zuckerman’s, and Eysenck’s dimensions of anterior cingulate activity is related to harm avoidance. personality. Pers. Individ. Differ. 21, 283–285. Neuroreport 20 (1), 19–23. van den Burg, E.H., Neumann, I.D., 2011. Bridging the gap Yassa, M.A., Hazlett, R.L., Stark, C.E., Hoehn-Saric, R., 2012. between GPCR activation and behaviour: oxytocin and Functional MRI of the amygdala and bed nucleus of the stria prolactin signalling in the hypothalamus. J. Mol. Neurosci. 43 terminalis during conditions of uncertainty in generalized (2), 200–208. anxiety disorder. J. Psychiatr. Res. 46 (8), 1045–1052. Zhang, L., Hernandez, V.S., Liu, B., Medina, M.P., Nava-Kopp, A.T., Yoshida, M., Takayanagi, Y., Inoue, K., Kimura, T., Young, L.J., Irles, C., et al., 2012. Hypothalamic vasopressin system Onaka, T., et al., 2009. Evidence that oxytocin exerts anxiolytic regulation by maternal separation: its impact on anxiety in effects via oxytocin receptor expressed in serotonergic rats. Neuroscience 215, 135–148. neurons in mice. J. Neurosci. 29 (7), 2259–2271. Zheng, D.J., Larsson, B., Phelps, S.M., Ophir, A.G., 2013. Female Young, L.J., 2002. The neurobiology of social recognition, alternative mating tactics, reproductive success and approach, and avoidance. Biol. Psychiatry 51 (1), 18–26. nonapeptide receptor expression in the social decision- Young, L.J., 2009. Being human: love: neuroscience reveals all. making network. Behav. Brain. Res. 246, 139–147. Nature 457 (7226), 148. Zheng, J.J., Li, S.J., Zhang, X.D., Miao, W.Y., Zhang, D., Yao, H., Young, L.J., Wang, Z., 2004. The neurobiology of pair bonding. Nat. et al., 2014. Oxytocin mediates early experience-dependent Neurosci. 7 (10), 1048–1054. cross-modal plasticity in the sensory cortices. Nature Young, S.N., Anderson, G.N., 2010. Bioanalytical inaccuracy: a neurosci.http://dxdoi.org/10.1038/nn.3634 (published threat to the integrity and efficiency of research. J. Psychiatry online 26 January 2014). Neurosci.: JPN 35 (1), 3–6. brain research 1580 (2014) 151–159

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Reading the mind in the infant eyes: Paradoxical effects of oxytocin on neural activity and emotion recognition in watching pictures of infant faces

Alexandra Voorthuisa,b, Madelon M.E. Riema,b, Marinus H. Van IJzendoorna,b,n, Marian J. Bakermans-Kranenburga,b aCentre for Child and Family Studies, Leiden University, The Netherlands bLeiden Institute for Brain and Cognition (LIBC), Leiden University, Leiden, The Netherlands article info abstract

Article history: The neuropeptide oxytocin facilitates parental caregiving and is involved in the processing Accepted 27 October 2013 of infant vocal cues. In this randomized-controlled trial with functional magnetic Available online 31 October 2013 resonance imaging we examined the influence of intranasally administered oxytocin on Keywords: neural activity during emotion recognition in infant faces. Blood oxygenation level Oxytocin dependent (BOLD) responses during emotion recognition were measured in 50 women Inferior frontal gyrus who were administered 16 IU of oxytocin or a placebo. Participants performed an adapted Superior temporal gyrus version of the Infant Facial Expressions of Emotions from Looking at Pictures (IFEEL Middle temporal gyrus pictures), a task that has been developed to assess the perception and interpretation of Infant facial expressions infants' facial expressions. Experimentally induced oxytocin levels increased activation in fMRI the inferior frontal gyrus (IFG), the middle temporal gyrus (MTG) and the superior temporal gyrus (STG). However, oxytocin decreased performance on the IFEEL picture task. Our findings suggest that oxytocin enhances processing of facial cues of the emotional state of infants on a neural level, but at the same time it may decrease the correct interpretation of infants' facial expressions on a behavior level. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2013 Elsevier B.V. All rights reserved.

1. Introduction involved in the processing of infant vocal cues (Riem et al., 2011, 2012b). Moreover, oxytocin has been shown to facilitate Before the onset of speech infants have vocalizations and mental state reasoning, defined as the ability to infer others' facial expressions at their disposal to communicate with the mental states, thought, feelings and intentions (Domes et al., caregiver. The correct interpretation of these infant signals is 2007b; Riem et al., submitted for publication). However, the essential for the caregiver in order to respond appropriately. aforementioned studies focus on infant vocal expressions and Previous studies have shown that oxytocin, a neuropeptide adult facial expressions. It is not yet clear whether oxytocin that is involved in social affiliation (Carter, 1998), is also plays a role in the interpretation of infant facial expressions.

nCorresponding author at: Leiden University, Centre for Child and Family Studies, PO Box 9555, 2300 RB Leiden, The Netherlands. Fax: 31 71 5273945. þ E-mail address: [email protected] (M.H. Van IJzendoorn).

0006-8993/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.brainres.2013.10.051 152 brain research 1580 (2014) 151–159

The current study is the first randomized controlled trial The orbitofrontal cortex (OFC) has also been shown to be examining the influence of intranasally administered oxyto- activated during the observation of infant faces (Rilling, 2013). cin on the neural mechanisms underlying emotion recogni- The OFC shows a very specific, rapid response to infant faces, tion in infant faces. We examined the neural response to and it has been suggested that this may be the potential brain infants' facial emotions with functional magnetic resonance basis for the “innate releasing mechanisms” for nurturing of imaging (fMRI). infants as described by Lorenz (Kringelbach et al., 2008). Infant facial expressions provide important cues about an Intranasal oxytocin has been shown to reduce activation infant's emotional state and are used by parents in concert and connectivity of the globus pallidus with reward- and with infant vocalizations and contextual information to attachment-related brain areas in fathers who were exposed interpret the needs of their infant. The “Infant Facial Expres- to pictures of their own child (Wittfoth-Schardt et al., 2012). It sions of Emotions from Looking at Pictures” (IFEEL Pictures; is as yet unknown how oxytocin influences neural responses Emde et al., 1993) have been developed to assess the percep- to infant faces in women. Moreover, it is unclear whether tion and interpretation of infants' facial expressions (e.g., intranasal oxytocin facilitates the ability to interpret infant Siddiqui et al., 2000). It consists of 30 pictures of infant faces facial expressions. and requires participants to describe in one word the emo- In this study, we examined the influence of intranasally tional expression of the infant. Several studies have shown administered oxytocin on neural activation during the IFEEL that the test can be used to study individual differences in the pictures task with fMRI in female adults. To our knowledge, interpretation of infant emotions. For example, research this study is the first randomized controlled trial examining using the IFEEL pictures has shown that abusive fathers the influence of intranasally administered oxytocin on the perceive infant emotions in a more negative light than non- neural mechanisms underlying emotion recognition in infant abusive fathers (Francis and Wolfe, 2008) and that neglecting faces. We were especially interested in effects of oxytocin on mothers are more inaccurate in labeling infant emotions the anterior insula, the inferior frontal gyrus pars opercularis (Hildyard and Wolfe, 2007). (IFG), and the orbitofrontal cortex (OFC), since previous Oxytocin is a neuropeptide that plays an important role in studies have shown that these regions play an important sensitive parenting. Previous research has shown that higher role in the perception of infant stimuli and can be affected by levels of maternal oxytocin across pregnancy predict higher intranasal oxytocin (Riem et al., 2011, 2012b; Strathearn et al., quality of postpartum maternal behavior (Feldman et al., 2009; Wittfoth-Schardt et al., 2012). In addition, we examined 2007) and that oxytocin has stress-reducing effects in breast- effects of intranasal oxytocin on the anterior superior tem- feeding mothers (Heinrichs et al., 2001, 2002). In addition, poral gyrus (STG) and the middle temporal gyrus temporo intranasal administration of oxytocin stimulates a range of occipital part (MTG), because these regions have been shown social behaviors (for a meta-analysis, see Van IJzendoorn and to be activated during emotion recognition in adult faces Bakermans-Kranenburg, 2012), such as trust (Kosfeld et al., (Adams et al., 2009; Pincus et al., 2010; Riem et al., submitted 2005) and empathy (Bartz et al., 2010). Domes et al. (2007b) for publication). Furthermore, since previous studies have showed that intranasal oxytocin also facilitates mindreading, found moderation of oxytocin effects by early caregiving assessed with the Reading the Mind in the Eyes Task (RMET; experiences (Bakermans-Kranenburg and Van IJzendoorn, Baron-Cohen et al., 1997), a test that requires individuals to 2013), we examined the potential moderating influence of infer mental states by looking at photographs of the eye experienced parenting on the effects of oxytocin. Finally, we region of adult faces. This is in line with other studies expected that intranasal oxytocin would increase perfor- showing that oxytocin stimulates emotion recognition in mance on the IFEEL picture task. adult faces (Bartz et al., 2010; Marsh et al., 2010), possibly by enhancing activation in empathy-related brain regions (Riem et al., submitted for publication). 2. Results FMRI studies measuring brain activity by detecting asso- ciated changes in blood flow have shown that intranasal In order to identify brain regions involved in reasoning about oxytocin facilitates sensitive responding to infant vocalizations the emotional state of infants, we contrasted the emotional by decreasing activation in the amygdala (Riem et al., 2012a), state condition with the gender condition. The whole brain abrainregionimplicatedintheexperienceoffearanddisgust analysis revealed three clusters in the placebo group with (LeDoux, 2000), and by increasing activation in the insula and peak voxels in the left middle temporal gyrus and in the inferior frontal gyrus (IFG) (Riem et al., 2011), brain regions bilateral inferior frontal gyrus (see Table 1). The pattern of involved in empathy, emotion recognition, and the processing activation included the bilateral paracingulate gyrus, the of facial emotional expressions (Bernhardt and Singer, 2012; temporal poles, the occipital poles, the bilateral occipital Carr et al., 2003; Keuken et al., 2011; Liakakis et al., 2011). Other cortex, the left putamen, the left insula, the right orbitofrontal studies also found activation in the anterior insula and IFG cortex, the bilateral fusiform gyri and the left thalamus (see during the presentation of pictures of infant faces (for a review Fig. 1). see Rilling, 2013). Insula and IFG activation may allow parents to To examine whether oxytocin affected brain activity internally simulate what a child is feeling. Indeed, in a previous during emotional state reasoning we contrasted the oxyto- study we found that intranasal oxytocin enhances activation in cin group with the placebo group (Oxytocin Emotion4Gender4 the insula during the Reading the Mind in the Eyes Task, Placebo Emotion4Gender and Oxytocin Emotion4GenderoPlacebo thereby facilitating the interpretation of adults' emotions Emotion4Gender). In the whole brain analysis no significant (Riem et al., submitted for publication). differences in brain activity were found between the brain research 1580 (2014) 151–159 153

Table 1 – MNI coordinates, cluster size, and Z-max values for significantly activated clusters.

Contrast Brain region MNI coordinates Cluster size Peak Z

xyz

Placebo Emotional state4Gender L Middle Temporal Gyrus 56 48 6 14,702 6.42 À À L Inferior Frontal Gyrus 54 16 22 11,346 7.06 À R Inferior Frontal Gyrus 56 26 2 1703 5.58 L Superior Temporal Gyrus 48 0 22 20 2.70a À À

Oxytocin Emotional state4Gender R Middle Temporal Gyrus 52 46 6 10,561 6.26 À L Inferior Frontal Gyrus 52 14 18 9472 7.63 À L Middle Temporal Gyrus 52 44 2 6043 7.09 À À L Superior Frontal Gyrus 4 14 58 1650 6.58 À L Superior Temporal Gyrus 48 0 22 147 5.70a À À

po.05, corrected by whole brain cluster threshold (Z42.3); use of oral contraceptives and menstrual cycle included as confound regressors in the model. a Region of interest analysis, po.05, corrected by cluster threshold (Z42.3).

Fig. 1 – Significant activation during the emotional state reasoning condition compared to the gender condition in the placebo group (the left sight of the brain corresponds with the right hemisphere and vice versa). Statistical images were thresholded with clusters determined by Z42.3 and a cluster-corrected significance threshold of po.05.

oxytocin and the placebo group. Region of interest analyses treatment (oxytocin/placebo) as a between-subject factor were conducted to test for significant differences in activa- showed a significant interaction between time and treatment tion during emotional state reasoning between the oxytocin F(1, 48) 12.28, p .001. Further analyses showed that the ¼ ¼ and placebo group in the a priori specified regions of oxytocin level significantly increased from before (M 6.38, ¼ interest. Oxytocin significantly increased activation in the SD 4.96) to after administration (M 822.11, SD 1163.16) ¼ ¼ ¼ left superior temporal gyrus (cluster size 60, peak Z 2.87, in the oxytocin group F(1, 24) 12.27, p .002) whereas the ¼ ¼ ¼ ¼ p .0021, MNI coordinates (mm) 52, 0, 14), the left oxytocin level did not significantly differ before (M 5.02, ¼ ¼À À ¼ inferior frontal gyrus (cluster size 24, peak Z 2.64, SD 2.81) and after placebo administration (M 4.63, ¼ ¼ ¼ ¼ p .0041, MNI coordinates (mm) 54, 12, 14) and the left SD 2.01) in the placebo group F(1, 24) .54, p .47. ¼ ¼À ¼ ¼ ¼ middle temporal gyrus (cluster size 47, peak Z 2.71, With regard to the performance on the IFEEL picture task, ¼ ¼ p .0034, MNI coordinates (mm) 46, 54, 12) (see Fig. 2). we found no relation between basal oxytocin levels and ¼ ¼À À Region of interest analyses with the right superior temporal performance (r .18, p .39) in the placebo group. The ¼À ¼ gyrus, the right inferior frontal gyrus, the right middle oxytocin group was less correct in recognizing the emotions temporal gyrus, the bilateral orbitofrontal cortex, and the (M 20.72, SD 2.49) than the placebo group (M 22.00, ¼ ¼ ¼ bilateral insula did not reveal significant differences SD 1.89), t(48) 2.045,p .046. Further inspection of the ¼ ¼À ¼ between the oxytocin and the placebo group. We examined data was done by using a repeated measure analysis includ- the potential moderating influence of the difficulty of the ing the difficulty levels of the items (easy, medium and pictures (ambiguity of the emotion) and participants' experi- difficult). Results showed a main effect of oxytocin (F(1, 48) ¼ ences with maternal rejection during childhood. Results 4.18, po.05), but no main effect of difficulty level (F(2, 47) .84, ¼ were not significant. p .44), and no interaction of oxytocin with difficulty level ¼ There were no differences in basal levels of oxytocin between (F(2, 47) .37, p .69). The basal level of oxytocin did not ¼ ¼ the oxytocin (M 6.38, SD 4.96) and placebo group (M 5.02, significantly add to the explained variance in neural activa- ¼ ¼ ¼ SD 2.81), t(48) 1.19, p .24. A repeated measure analysis with tion (IFG beta .210, p .148; STG beta .034, p .821, MTG ¼ ¼ ¼ ¼ ¼ ¼ ¼ time (before/after administration) as a within-subject factor and beta .145, p .325). There were no significant associations À ¼ 154 brain research 1580 (2014) 151–159

Fig. 2 – (a) Oxytocin effects on the left IFG, the left MTG and the left STG. Region of interest analysis, po.05, corrected by cluster threshold Z42.3 and (b) Z values (M, SE) of the left IFG, left MTG and left STG activation for the emotion state, gender and emotion4gender condition. between the number of correct responses and increased brain activation in the IFG (r .08, p .57), the STG (r .10, recognition in adult faces (e.g., Adams et al., 2009; Baron- ¼À ¼ ¼À p .48) and the MTG (r .02, p .89). Cohen et al., 1997; Critchley et al., 2000) and neural responses ¼ ¼À ¼ to infant faces (e.g., Leibenluft et al., 2004; Lenzi et al., 2009; Strathearn et al., 2008) also found activation of the IFG, the 3. Discussion STG and the MTG. The MTG is part of the ventral visual processing pathway, which is involved in the conscious and We found that intranasally administered oxytocin enhanced explicit appraisal of facial expressions (Critchley et al., 2000). activation in the left IFG, MTG and STG during emotion Increased activation in the MTG and the STG might indicate a recognition in infant faces compared to gender identification. more conscious evaluation of facial expressions and This is in line with studies showing neural activation in these enhanced processing of facial emotions. Furthermore, we areas during emotion recognition in adult faces (Adams et al., found that oxytocin had no effect on the insula, possibly 2009; Critchley et al., 2000), and with studies showing that because the insula is mostly involved in the perception of these regions are affected by intranasal oxytocin (Pincus negative emotions (Lang et al., 2011). One explanation for the et al., 2010; Riem et al., 2011). At the same time, oxytocin absence of oxytocin effects on the OFC may be that the OFC increased the number of mistakes in recognizing infant response to infant faces is too rapid and transient to be emotions in the IFEEL picture task. detected using fMRI (Kringelbach et al., 2008) and because The IFG is considered an important part of the human scanning of the OFC is prone to artifacts due to its proximity mirror neuron system (Gallese and Goldman, 1998) and has to air-filled sinuses (Kringelbach et al., 2004). shown to be active during imitation and observation of In the current study we found that oxytocin primarily emotional facial expressions of infants (Lenzi et al., 2009). increased activation in the left hemisphere. Similarly, in a Activation in the mirror neuron system facilitates the under- previous study we found that oxytocin effects on neural standing of others' thoughts and feelings (Carr et al., 2003). activation during the RMET were restricted to the left hemi- Previous studies examining neural activation during emotion sphere (Riem et al., submitted for publication). There is some brain research 1580 (2014) 151–159 155

evidence that suggests that positive facial expressions are The limitations of our study should be acknowledged. One processed primarily in the left hemisphere, in contrast to limitation is that we used a between-subjects design to study negative facial expressions, which are processed in the right the effects of oxytocin, which implies the risk of pre-existing hemisphere (Canli et al., 1998). This is consistent with the differences between the oxytocin and the placebo group. traditional view that the right hemisphere is specialized in Randomization and double-blind application have decreased handling negative emotions, whereas the left hemisphere this risk substantially. In addition, meta-analytic findings specializes on handling positive emotions (Silberman and indicate that studies using between- or within-subject Weingartner, 1986). However, recent research indicates that design to examine intranasal oxytocin influences show no differences in left and right hemispheric activation may significant differences in effect size (Van IJzendoorn and relate to motivational direction rather than emotional Bakermans-Kranenburg, 2012). Although we did not expect valence. More specifically, greater activity in left frontal and pre-existing differences in emotion recognition ability and anterior temporal regions has been shown to reflect approach tested for basal oxytocin levels between the two groups, there motivation, whereas greater right activity reflects greater may have been differences between individuals in emotion withdrawal motivation (Davidson, 1992; Harmon-Jones recognition ability as related to genetic differences, e.g. in the et al., 2010). Our finding that the effects of oxytocin are oxytocin receptor OXTR gene polymorphism (Bartz et al., restricted to the left hemisphere might therefore indicate 2011). We found however no difference in basal oxytocin that oxytocin enhances approach-related behavior. Indeed, levels, no associations between basal oxytocin and emotion previous studies have found that oxytocin enhances heart recognition, and in a meta-analysis differences in OXTR rate variability, a psychophysiological marker of approach- rs53576 and rs2254298 were not related to social-emotional related motivation (Kemp et al., 2012) and that it is involved behavior (Bakermans-Kranenburg and Van IJzendoorn, 2013). in our motivation to be with others (for a review see Gordon The inclusion of a sample of females without children is et al., 2011). another limitation of this study. Earlier studies showed that In previous research with the IFEEL pictures parents showed oxytocin affects brain activity when looking at faces with an improvement in their ability to appropriately interpret emotional expressions differently in males and females infants' emotional cues after an intervention aimed at enhan- (Domes et al., 2007a, 2010). Although the generalizability of cing caregiving quality (Constantino et al., 2001). Moreover, our findings may be limited, examining oxytocin effects in neglectful mothers were less accurate in recognizing infant women fills an important gap as long as most previous emotions than non-neglecting mothers (Hildyard and Wolfe, studies focused on the effects of oxytocin in men (Bos et al., 2007). In contrast with our expectations, we found that 2012). In our sample, 70% of the participants used oral oxytocin decreased emotion recognition performance. The contraceptives. As contraceptives interfere with the endo- effect was independent of the difficulty level of the items. genous steroid metabolism and steroids may interact with Earlier studies found more pronounced oxytocin effects for intranasally administered oxytocin, we took the use of oral difficult compared to easy items (Domes et al., 2007b; Marsh contraceptives into account as a covariate. The covariate did et al., 2010). These studies examined emotion recognition in not affect our results. Future studies may examine the adult faces while we examined emotion recognition in infant possible interaction of intranasal oxytocin and the use of faces. It may be that the influence of oxytocin on emotion oral contraceptives in larger samples. Furthermore, we found recognition with adult faces shows more variability depend- no moderating effect of early childhood experiences on ing on ambiguity of the emotion. Further research comparing neural activation, and further research is needed to clarify adult faces and infant faces with varying levels of ambiguity whether and how the effects of oxytocin vary depending on of the expressed emotions may shed light on this issue. To individual factors. our surprise we found that oxytocin enhanced neural activity We used intranasal oxytocin administration to examine in empathy-related brain regions during emotion recognition, changes in brain activation and behavior. It has been reported while it decreased performance in emotion recognition. that related neuropeptides such as arginine vasopressin We found no significant relation between brain activation (AVP) can cross the blood brain barrier (BBB) and enter the and performance. This is in line with earlier studies in which brain directly from the nasal capillaries via the olfactory bulb RMET performance was not related to brain activity during neurons (Born et al., 2002). Other routes by which these the RMET task (Riem et al., submitted for publication, but see neuropeptides may reach the brain are for example via the Adams et al., 2009) and neural responses to infant crying did ventricular CSF followed by entering the extra-cellular space not predict behavioral responses to crying (Riem et al., 2012a). of the brain (Born et al., 2002), or via trigeminal nerve Thus, there seems to be a transmission gap between task- branches into the brain (Guastella et al., 2012). It is likely related brain activity and actual behavior. Berkman and Falk that multiple routes operate independently of each other (2013) recently discussed the brain-as-predictor approach (Neumann et al., 2013) and that the uptake of intranasally with neural measures as independent variables in models administered oxytocin by one or more of these routes shows with observed “real-world” outcomes as dependent variables. inter-individual variation in relative contribution in the net They describe established brain-behavior relations including uptake (Guastella et al., 2012). neural signals that predict skill acquisition or cognitive decline. A recent study with rodents showed that oxytocin levels in Future research may implement the brain-as-predictor approach both the amygdala and hippocampus were increased after in the investigation of parent–infant interactions and examine intranasal oxytocin administration, and that this increase in the relation between task-related brain activity and actual the brain was paralleled by changes in plasma levels of caregiving behavior. oxytocin (Neumann et al., 2013), but it is unclear whether 156 brain research 1580 (2014) 151–159

these findings generalize to humans. However, the results of randomly assigned to the oxytocin (n 25) or the placebo ¼ our studies and those of others (for a review see Rilling, 2013) condition (n 25). The mean age of the participants was 19.66 ¼ show that intranasal administered oxytocin does have effects years (SD 1.45, range 18–27) and the majority (70%) used oral ¼ on neural activity and behavior (for a meta-analysis, see Van contraceptives. Permission for this study was obtained from IJzendoorn and Bakermans-Kranenburg, 2012). Moreover, ele- the Ethics Committee of the Institute of Education and Child vated levels of oxytocin in saliva have been demonstrated up Studies of Leiden University and of the Leiden University to seven hours after nasal administration (Van IJzendoorn Medical Centre. et al., 2012). The persistence of intranasal oxytocin levels in saliva has been suggested to be caused by a feed forward 4.2. Procedure mechanism in which treatment with intranasal oxytocin might lead to an increase in the production of endogenous The majority of the participants (82%) were invited in the oxytocin (Van IJzendoorn et al., 2012). These results support luteal phase of their menstrual cycle, which was assessed the validity of the procedure, but further research is needed using self-report. It was not possible to determine menstrual to establish how exactly oxytocin enters the brain and in phase for two participants because of use of Mirena intrau- what amounts it reaches the receptors in which regions of terine device. Participants took six puffs of nasal spray the human brain. containing oxytocin (16 IU total) or placebo spray approxi- In sum, this is the first randomized controlled trial exam- mately 50–60 minutes before the start of the fMRI data ining the influence of intranasally administered oxytocin on acquisition for the IFEEL pictures task. The administration the neural mechanisms underlying emotion recognition in of the spray was done under supervision of the experimenter infant faces. We found that oxytocin enhanced activation in and drug administration was double-blind. After administra- brain regions implicated in empathy and emotion understand- tion of the nasal spray, the participants completed some ing when individuals infer the emotional state of infants with practice trials outside the scanner to become familiar with various facial expressions, which may facilitate the interpreta- the task. Salivary oxytocin levels have been shown to remain tion of the needs of infants and the selection of a sensitive elevated for more than seven hours after administration of caregiving response. Surprisingly, we found that oxytocin nasal spray containing 16 IU of oxytocin (Van IJzendoorn lowered performance in emotion recognition in infants and et al., 2012) and effects of nasal spray with 16 IU of oxytocin that performance and brain activation were not related. The on social behavior and neural activity have been reported in paradoxical effects of oxytocin with enhanced neural activa- previous studies (Riem et al., 2011). tion of empathy-related brain regions on the one hand and lower performance on emotion recognition of infant faces on 4.3. Experimental task the other hand should be focus of further research. Participants performed an adapted version of the IFEEL pictures task (Emde et al., 1993). Participants viewed 30 4. Experimental procedure photographs of infant faces in two different conditions, resulting in a total of 60 stimulus presentations. In the first 4.1. Participants condition, two emotion words accompanied each photograph and participants were asked to select the word that best A total of 343 female undergraduate students from the described what the infant in the photograph was thinking or departments of education and child studies, and psychology feeling. The selection of the two emotion words for each at Leiden University participated in the first phase of the picture was based on scores of a reference group of 100 study. In this phase, the participants completed online ques- mothers of infants (Emde et al., 1993). For each picture, two tionnaires on experienced parenting and some demographic emotion words were selected as the emotion words in the details. One participant was excluded due to random current fMRI study. The emotion that was more often chosen responses and five females with children of their own were by the reference group (Emde et al., 1993) was considered the excluded to avoid effects of parental experiences on the correct response, and the emotion that was less often chosen outcomes. One hundred eighty-six nulliparous females parti- was considered the incorrect response. Correct and incorrect cipated in the second phase of the study in which behavioral emotions were presented counterbalanced on the left and responses to infant crying were examined. Fifty-four partici- right side of the screen. In the control condition, the partici- pants with scores ranging from low to high on an experienced pants were asked to indicate whether the infant was a boy parenting questionnaire were selected to participate in the or a girl. Participants were asked to indicate the child's third phase of the study. The third phase consisted of a emotional state or to indicate the gender of the child by computer game designed to study prosocial helping behavior pressing buttons with the right hand. The photographs were towards an excluded adult (without an fMRI component, see presented in blocks of 5 trials. Each photograph was pre- Riem et al., 2013), and the current fMRI study. Participants sented for 5 s (interstimulus time 1–2.5 s). To compare the were screened for MRI contraindications, psychiatric or responses to the relatively easy (unambiguous) and more neurological disorders, hearing problems, pregnancy, and difficult (ambiguous) pictures, we distinguished three groups: alcohol and drug abuse. Due to excessive head movement easy, medium and difficult. For each picture we calculated a during fMRI scanning, four participants were excluded from difference score based on differences in frequencies of the the current analyses, resulting in a total sample of 50 alternatives in the reference group. For example, when for a nulliparous female participants. The participants were specific picture 94% of the reference group chose “content” brain research 1580 (2014) 151–159 157

and 5% chose “passive”, the difference score was (94–5 )89. and we tested for group differences using two-sample t-tests ¼ Ten pictures (with difference scores 31–90) were classified as on these contrasts with the oxytocin versus placebo group easy, ten as medium (difference scores 12–30) and ten as comparison (Oxytocin4Placebo and OxytocinoPlacebo). difficult (difference scores 1–11). Menstrual cycle phase and We included menstrual cycle (follicular or luteal phase) and oral contraceptive use had no influence on performance on use of oral contraceptives as confound regressors in the model the IFEEL pictures task. in the analyses of the group means and group differences. Menstrual cycle and use of oral contraceptives showed no 4.4. Maternal love withdrawal. effects on neural activity during the task. The statistical images were thresholded using clusters determined by Maternal use of love withdrawal was assessed using an Z42.3 and a cluster corrected significance threshold of po.05. 11-item questionnaire (Huffmeijer et al., 2011) based on the Region of Interest (ROI) analyses were conducted with a Child's Report of Parental Behavior Inventory (CRPBI; Beyers priori specified regions of interest based on previous fMRI and Goossens, 2003; Schludermann and Schludermann, 1970) studies (e.g., Adams et al., 2009; Castelli et al., 2010; Moor and the Parental Discipline Questionnaire (PDQ; Hoffman and et al., 2012; O'Doherty et al., 2003). These regions were the Saltzste, 1967; Patrick and Gibbs, 2007). Each item (e.g. My anterior insula, the inferior frontal gyrus pars opercularis mother was a person who is less friendly with me, if I do not see (IFG), the anterior superior temporal gyrus (STG), the orbito- things her way) was rated on a 5-point scale ranging from not frontal cortex (OFC) and the middle temporal gyrus temporo applicable to fully applicable. Due to a technical error one of the occipital part (MTG) (475% probability, anatomically defined items was missing in some questionnaires. The score for this using the Harvard–Oxford cortical atlas, http://fsl.fmrib.ox.ac. item was substituted with the mean score of the other items. uk/fsl/fslwiki/Atlases). ROI analyses were limited to these Internal consistency of the scale was high (Cronbach's α .92). regions, applying the same statistical threshold as for the ¼ whole brain analyses, but correcting only for the size of ROI 4.5. fMRI data acquisition and analysis volumes. The correction for multiple testing was done for all three ROIs separately. Mean Z-values for significantly acti- Scanning was performed with a standard whole-head coil on vated voxels in the left IFG, the left STG and the left MTG a 3-T Philips Achieva TX MRI system (Philips Medical Sys- were calculated for visualization purposes. tems, Best, the Netherlands) in the Leiden University Medical Center. First, a T1-weighted anatomical scan was acquired (flip angle 81, 140 slices, voxel size.875 875 1.2 mm). For ¼   fMRI, a total of 186 T2n-weighted whole-brain echoplanar images were acquired (repetition time 2.2 s; echo time Acknowledgments ¼ ¼ 30 ms, flip angle 801, 38 transverse slices, voxel size ¼ 2.75 2.75 2.75 mm ( 10% interslice gap)). All anatomical MJBK and MHvIJ were supported by awards from the   þ fi scans were examined by a radiologist in accordance with Netherlands Organization for Scienti cResearch(NWO) Leiden University Medical Center policy. No anomalies (MJBK: VICI Grant; MHvIJ: SPINOZA prize). were found. Data analysis was carried out using FEAT (FMRI Expert references Analysis Tool) version 5.98, part of FSL (Smith et al., 2004). The following pre-statistics processing was applied: motion correction (Jenkinson et al., 2002), non-brain removal (Smith, Adams, R.B., Rule, N.O., Franklin, R.G., Wang, E., Stevenson, M.T., 2002), spatial smoothing using a Gaussian kernel of full- Yoshikawa, S., Nomura, M., Sato, W., Kveraga, K., Ambady, N., 2009. Cross-cultural reading the mind in the eyes: an fMRI width-at-half-maximum 8.0 mm, and high-pass temporal investigation. J. Cognit. Neurosci. 22, 97–108. filtering (highpass filter cutoff 100.0 s). Functional scans ¼ Bakermans-Kranenburg, M.J., Van IJzendoorn, M.H., 2013. Sniffing were registered to the high-resolution EPI-images, which around oxytocin: review and meta-analyses of trials in were registered to the T1-weighted images, which were healthy and clinical groups with implications for registered to standard space (Jenkinson et al., 2002). pharmacotherapy. Transl. Psychiatry 3, e258. In native space, functional activation was examined using Baron-Cohen, S., Jolliffe, T., Mortimore, C., Robertson, M., 1997. general linear model analysis. Each condition (emotional Another advanced test of theory of mind: evidence from very high functioning adults with autism or Asperger syndrome. state, gender) was modeled separately as a square-wave J. Child Psychol. Psychiatry 38, 813–822. function. Each predictor was then convolved with a double Bartz, J.A., Zaki, J., Bolger, N., Hollander, E., Ludwig, N.N., gamma hemodynamic response function and its temporal Kolevzon, A., Ochsner, K.N., 2010. Oxytocin selectively derivative was added to the model, giving four regressors. To improves empathic accuracy. Psychol. Sci. 21, 1426–1428. be able to examine brain regions activated during the emo- Bartz, J.A., Zaki, J., Bolger, N., Ochsner, K.N., 2011. Social effects of tional state attribution we contrasted the emotional state oxytocin in humans: context and person matter. Trends attribution condition with the gender condition (emotion4 Cognit. Sci. 15 (7), 301–309, http://dx.doi.org/10.1016/j. tics.2011.05.002. gender). Berkman, E.T., Falk, E.B., 2013. Beyond brain mapping: using The first-level contrast images and the corresponding neural measures to predict real-world outcomes. Curr. Dir. variance images were transformed to standard space and Psychol. Sci. 22, 45–50. submitted to second-level mixed-effects group whole brain Bernhardt, B.C., Singer, T., 2012. The neural basis of empathy. analyses. Group means were tested using one-sample t-tests Annu. Rev. Neurosci. vol 35 (35), 1–23. 158 brain research 1580 (2014) 151–159

Beyers, W., Goossens, L., 2003. Psychological separation and Harmon-Jones, E., Gable, P.A., Peterson, C.K., 2010. The role of adjustment to university: moderating effects of gender, age, asymmetric frontal cortical activity in emotion-related and perceived parenting style. J. Adolescent Res. 18, 363–382. phenomena: a review and update. Biol. Psychol. 84, 451–462. Born, J., Lange, T., Kern, W., McGregor, G.P., Bickel, U., Fehm, H.L., Heinrichs, M., Meinlschmidt, G., Neumann, I., Wagner, S., 2002. Sniffing neuropeptides: a transnasal approach to the Kirschbaum, C., Ehlert, U., Hellhammer, D.H., 2001. Effects of human brain. Nat. Neurosci. 5, 514–516. suckling on hypothalamic-pituitary-adrenal axis responses to Bos, P.A., Panksepp, J., Bluthe, R.M., van Honk, J., 2012. Acute psychosocial stress in postpartum lactating women. J. Clin. effects of steroid hormones and neuropeptides on human Endocrinol. Metab. 86, 4798–4804. social-emotional behavior: a review of single administration Heinrichs, M., Neumann, I., Ehlert, U., 2002. Lactation and stress: studies. Front. Neuroendocrinol. 33, 17–35. protective effects of breast-feeding in humans. Stress 5, Canli, T., Desmond, J.E., Zhao, Z., Glover, G., Gabrieli, J.D., 1998. 195–203. Hemispheric asymmetry for emotional stimuli detected with Hildyard, K., Wolfe, D., 2007. Cognitive processes associated with fMRI. NeuroReport 9, 3233–3239. child neglect. Child Abuse Negl. 31, 895–907. Carr, L., Iacoboni, M., Dubeau, M.C., Mazziotta, J.C., Lenzi, G.L., Hoffman, M.L., Saltzste, H.D., 1967. Parent discipline and the 2003. Neural mechanisms of empathy in humans: a relay from child's moral development. J. Pers. Soc. Psychol. 5, 45–57. neural systems for imitation to limbic areas. Proc. Natl. Acad. Huffmeijer, R., Tops, M., Alink, L.R.A., Bakermans-Kranenburg, Sci. USA 100, 5497–5502. M.J., Van IJzendoorn, M.H., 2011. Love withdrawal is related to Carter, C.S., 1998. Neuroendocrine perspectives on social heightened processing of faces with emotional expressions attachment and love. Psychoneuroendocrinology 23, 779–818. and incongruent emotional feedback: evidence from ERPs. Castelli, I., Baglio, F., Blasi, V., Alberoni, M., Falini, A., Liverta- Biol. Psychol. 86, 307–313. Sempio, O., Nemni, R., Marchetti, A., 2010. Effects of aging on Jenkinson, M., Bannister, P., Brady, M., Smith, S., 2002. Improved mindreading ability through the eyes: an fMRI study. optimization for the robust and accurate linear registration – Neuropsychologia 48, 2586 2594. and motion correction of brain images. NeuroImage 17, Constantino, J.N., Hashemi, N., Solis, E., Alon, T., Haley, S., 825–841. McClure, S., Carlson, V.K., 2001. Supplementation of urban Kemp, A.H., Quintana, D.S., Kuhnert, R.L., Griffiths, K., Hickie, I.B., home visitation with a series of group meetings for parents Guastella, A.J., 2012. Oxytocin increases heart rate variability and infants: results of a “real-world” randomized controlled in humans at rest: implications for social approach-related trial. Child Abuse Negl. 25, 1571–1581. motivation and capacity for social engagement. PLoS One 7, Critchley, H., Daly, E., Phillips, M., Brammer, M., Bullmore, E., e44014. Williams, S., Van Amelsvoort, T., Robertson, D., David, A., Keuken, M.C., Hardie, A., Dorn, B.T., Dev, S., Paulus, M.P., Jonas, Murphy, D., 2000. Explicit and implicit neural mechanisms for K.J., Van Den Wildenberg, W.P.M., Pineda, J.A., 2011. The role processing of social information from facial expressions: a of the left inferior frontal gyrus in social perception: an rTMS functional magnetic resonance imaging study. Hum. Brain study. Brain Res. 1383, 196–205. Mapp. 9, 93–105. Kosfeld, M., Heinrichs, M., Zak, P.J., Fischbacher, U., Fehr, E., 2005. Davidson, R.J., 1992. Anterior cerebral asymmetry and the nature Oxytocin increases trust in humans. Nature 435, 673–676. of emotion. Brain Cognit. 20, 125–151. Kringelbach, M.L., de Araujo, I.E., Rolls, E.T., 2004. Taste-related Domes, G., Heinrichs, M., Glascher, J., Buchel, C., Braus, D.F., activity in the human dorsolateral prefrontal cortex. Herpertz, S.C., 2007a. Oxytocin attenuates amygdala NeuroImage 21, 781–788. responses to emotional faces regardless of valence. Biol. Kringelbach, M.L., Lehtonen, A., Squire, S., Harvey, A.G., Craske, Psychiatry 62, 1187–1190. M.G., Holliday, I.E., Green, A.L., Aziz, T.Z., Hansen, P.C., Domes, G., Heinrichs, M., Michel, A., Berger, C., Herpertz, S.C., Cornelissen, P.L., Stein, A., 2008. A specific and rapid neural 2007b. Oxytocin improves “mind-reading” in humans. Biol. signature for parental instinct. PLoS One, 3. Psychiatry 61, 731–733. Domes, G., Lischke, A., Berger, C., Grossmann, A., Hauenstein, K., Lang, S., Yu, T., Markl, A., Muller, F., Kotchoubey, B., 2011. Hearing Heinrichs, M., Herpertz, S.C., 2010. Effects of intranasal others' pain: neural activity related to empathy. Cognit. Affect. – oxytocin on emotional face processing in women. Behav. Neurosci. 11, 386 395. Psychoneuroendocrinology 35, 83–93. LeDoux, J.E., 2000. Emotion circuits in the brain. Annu. Rev. Emde, R.D., Osofsky, J.D., Butterfield, P.M., 1993. The IFEEL Neurosci. 23, 155–184. Pictures, a New Instrument for Interpreting Emotions. Leibenluft, E., Gobbini, M.I., Harrison, T., Haxby, J.V., 2004. International Universities Press, Inc, Madison, WI. Mothers' neural activation in response to pictures of their Feldman, R., Weller, A., Zagoory-Sharon, O., Levine, A., 2007. children and other children. Biol. Psychiatry 56, 225–232. Evidence for a neuroendocrinological foundation of human Lenzi, D., Trentini, C., Pantano, P., Macaluso, E., Iacoboni, M., affiliation: plasma oxytocin levels across pregnancy and the Lenzi, G.L., Ammaniti, M., 2009. Neural basis of maternal postpartum period predict mother–infant bonding. Psychol. communication and emotional expression processing during Sci. 18, 965–970. infant preverbal stage. Cereb. Cortex 19, 1124–1133. Francis, K.J., Wolfe, D.A., 2008. Cognitive and emotional Liakakis, G., Nickel, J., Seitz, R.J., 2011. Diversity of the inferior differences between abusive and non-abusive fathers. Child frontal gyrus: a meta-analysis of neuroimaging studies. Abuse Negl. 32, 1127–1137. Behav. Brain Res. 225, 341–347. Gallese, V., Goldman, A., 1998. Mirror neurons and the simulation Marsh, A.A., Yu, H.H., Pine, D.S., Blair, R.J., 2010. Oxytocin theory of mind-reading. Trends Cognit. Sci. 2, 493–501. improves specific recognition of positive facial expressions. Gordon, I., Martin, C., Feldman, R., Leckman, J.F., 2011. Oxytocin Psychopharmacology (Berl) 209, 225–232. and social motivation. Dev. Cognit. Neurosci. 1, 471–493. Moor, B.G., Macks, Z.A., Guroglu, B., Rombouts, S.A., Molen, M.W., Guastella, A.J., Hickie, I.B., McGuinness, M.M., Otis, M., Woods, E.A., Crone, E.A., 2012. Neurodevelopmental changes of reading the Disinger, H.M., Chan, H.K., Chen, T.F., Banati, R.B., 2012. mind in the eyes. Soc. Cognit. Affect. Neurosci. 7, 44–52. Recommendations for the standardisation of oxytocin nasal Neumann, I.D., Maloumby, R., Beiderbeck, D.I., Lukas, M., administration and guidelines for its reporting in human Landgraf, R., 2013. Increased brain and plasma oxytocin after research. Psychoneuroendocrinology 38 (5) http://dx.doi.org/ nasal and peripheral administration in rats and mice. 10.1016/j.psyneuen.2012.11.019 (612-25). Psychoneuroendocrinology. brain research 1580 (2014) 151–159 159

O'Doherty, J., Winston, J., Critchley, H., Perrett, D., Burt, D.M., Schludermann, E., Schludermann, S., 1970. Replicability of factors Dolan, R.J., 2003. Beauty in a smile: the role of medial in Children's Report of Parent Behavior (CRPBI). J. Psychol. 76, orbitofrontal cortex in facial attractiveness. Neuropsychologia 239–249. 41, 147–155. Siddiqui, A., Eisemann, M., H#auaggl#auof, B., 2000. The stability Patrick, R.B., Gibbs, J.C., 2007. Parental expression of of maternal interpretation of infant's facial expressions disappointment: should it be a factor in Hoffman's model of during pre-and postnatal period and its relation to prenatal – parental discipline? J. Genet. Psychol. 168, 131 145. attachment. Early Child Dev. Care 162, 41–50. Pincus, D., Kose, S., Arana, A., Johnson, K., Morgan, P.S., Silberman, E.K., Weingartner, H., 1986. Hemispheric lateralization Borckardt, J., Herbsman, T., Hardaway, F., George, M.S., of functions related to emotion. Brain Cognit. 5, 322–353. Panksepp, J., Nahas, Z., 2010. Inverse effects of oxytocin on Smith, S.M., 2002. Fast robust automated brain extraction. Hum. attributing mental activity to others in depressed and healthy Brain Mapp. 17, 143–155. subjects: a double-blind placebo controlled FMRI study. Front. Psychiatry 1, 134. Smith, S.M., Jenkinson, M., Woolrich, M.W., Beckmann, C.F., Riem, M.M.E., Bakermans-Kranenburg, M.J., Huffmeijer, R., Van Behrens, T.E., Johansen-Berg, H., Bannister, P.R., De Luca, M., IJzendoorn, M.H., 2013. Does intranasal oxytocin promote Drobnjak, I., Flitney, D.E., Niazy, R.K., Saunders, J., Vickers, J., prosocial behavior to an excluded fellow player? A Zhang, Y., De Stefano, N., Brady, J.M., Matthews, P.M., 2004. randomized-controlled trial with Cyberball. Advances in functional and structural MR image analysis and Psychoneuroendocrinology 38, 1418–1425. implementation as FSL. NeuroImage 23 (Suppl 1), S208–219. Riem, M.M.E., Bakermans-Kranenburg, M.J., Pieper, S., Tops, M., Strathearn, L., Fonagy, P., Amico, J., Montague, P.R., 2009. Adult Boksem, M.A., Vermeiren, R.R., Van IJzendoorn, M.H., attachment predicts maternal brain and oxytocin response to Rombouts, S.A., 2011. Oxytocin modulates amygdala, insula, infant cues. Neuropsychopharmacology 34, 2655–2666. and inferior frontal gyrus responses to infant crying: a Strathearn, L., Li, J., Fonagy, P., Montague, P.R., 2008. What's in a randomized controlled trial. Biol. Psychiatry 70, 291–297. smile? Maternal brain responses to infant facial cues. Riem, M.M.E., Bakermans-Kranenburg, M.J., Van IJzendoorn, M.H., Pediatrics 122, 40–51. Out, D., Rombouts, S.A., 2012a. Attachment in the brain: adult Van IJzendoorn, M.H., Bakermans-Kranenburg, M.J., 2012. A sniff attachment representations predict amygdala and behavioral of trust: meta-analysis of the effects of intranasal oxytocin responses to infant crying. Attachment Hum. Dev. 14, 533–551. administration on face recognition, trust to in-group, and Riem, M.M.E., Bakermans-Kranenburg, M.J., Voorthuis, A., Van trust to out-group. Psychoneuroendocrinology 37, 438–443. IJzendoorn, M.H., 2013. Oxytocin effects on mind reading are Van IJzendoorn, M.H., Bhandari, R., Van der Veen, R., Grewen, moderated by experiences of parental love withdrawal: an K.M., Bakermans-Kranenburg, M.J., 2012. Elevated salivary fMRI study (submitted for publication). Riem, M.M.E., Van IJzendoorn, M.H., Tops, M., Boksem, M.A., levels of oxytocin persist more than 7 h after intranasal Rombouts, S.A., Bakermans-Kranenburg, M.J., 2012b. No administration. Front. Neurosci. 6, 174. laughing matter: intranasal oxytocin administration changes Wittfoth-Schardt, D., Grunding, J., Wittfoth, M., Lanfermann, H., functional brain connectivity during exposure to infant Heinrichs, M., Domes, G., Buchheim, A., Gündel, H., Waller, C., laughter. Neuropsychopharmacology 37, 1257–1266. 2012. Oxytocin modulates neural reactivity to children's faces Rilling, J.K., 2013. The neural and hormonal bases of human as a function of social salience. Neuropsychopharmacology parental care. Neuropsychologia 51, 731–747. 37, 1799–1807. brain research 1580 (2014) 102–123

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report Schizophrenia and alcohol dependence: Diverse clinical effects of oxytocin and their evolutionary origins

Cort A. Pedersen

Department of Psychiatry, CB# 7160, The University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7160, USA article info abstract

Article history: Beginning in 1979 with the first report that central administration of oxytocin stimulates Accepted 28 January 2014 maternal behavior in virgin rats, decades of animal research and more recent human Available online 5 February 2014 studies have demonstrated that oxytocin has many pro-social effects. These many findings suggest that oxytocin may be an effective treatment for social deficits that are hallmark Keywords: features of disorders such as autism and schizophrenia. Effects in preclinical animal models Oxytocin also imply that oxytocin may be an efficacious pharmacotherapy in a wide range of Schizophrenia psychiatric disorders including psychoses and addictions. To date, 3 small clinical trials Social cognition found that daily intranasal oxytocin treatment for 2–8 weeks significantly reduced psychotic Alcohol withdrawal symptoms in schizophrenia. Two of these trials also found improvement in social cognition Alcohol dependence or neurocognition, areas in which patients have significant deficiencies that do not respond Treatment to conventional antipsychotic treatment and contribute to disability. In another small trial, intranasal oxytocin potently blocked alcohol withdrawal. After reviewing the rationale for these trials, they are described in more detail. Questions are then asked followed by discussions of the large gaps in our knowledge about brain oxytocin systems in humans. The hope is to highlight important directions for future investigations of the role of oxytocin in the pathophysiology of psychotic disorders and addictions and to extend clinical research in these areas. Heretofore unrecognized roles for which oxytocin may have been selected during the evolution of placental mammalian maternal–infant and other social attachments are considered as possible origins of oxytocin antipsychotic and antiaddiction effects. This article is part of a Special Issue entitled Oxytocin and Social Behav. & 2014 Published by Elsevier B.V.

1. Evolutionary prologue is essential for the success of lactation, and stimulation of uterine contractions, which facilitates parturition following Chapter 2 (in this volume) discusses the evolution of oxytocin fetal development within the uterus. Chapter 2 also empha- (OT) and its central roles in unique aspects of placental sizes that OT was selected to activate avid and sustai- mammalian reproduction. These include milk ejection, which ned maternal behavior in parturient females, another novel

Abbreviations: OT, oxytocin E-mail address: [email protected] http://dx.doi.org/10.1016/j.brainres.2014.01.050 0006-8993 & 2014 Published by Elsevier B.V. brain research 1580 (2014) 102–123 103

reproductive strategy that was essential for the successful mates. OT could have been selected to play this role. As a evolution of placental mammals. It was hypothesized that by-product, OT may block tolerance formation to other strong the much higher quality maternal behavior exhibited by activators of reward pathways including addictive drugs. placental mammals required the evolution of new brain systems that could initiate and maintain a highly motivated social behavioral state, systems that OT activates at parturi- tion. OT (as well as vasopressin) appears to have been 2. Introduction selected to initiate other social attachments that evolved in placental mammals, such as monogamous pair-bonding. Three recent clinical trials have rapidly propelled OT into the Evidence was reviewed that OT may also have been selected forefront of schizophrenia research. All 3 trials found that to regulate other social and emotion behavior control sys- twice daily intranasal administration of OT for 2–8 weeks as tems that emerged during placental mammalian evolution to an adjunct to antipsychotic medication significantly reduced enhance reproductive success. For example, OT has been positive, negative and general symptoms in patients with strongly implicated in reducing anxiety and fear responses schizophrenia (Feifel et al., 2010; Pedersen et al., 2011; during parturition and lactation which enabled successful Modabbernia et al., 2013). Furthermore, OT treatment but activation of offspring caretaking in parturient mothers, not placebo improved performance on tests of social cogni- especially primiparous mothers in whom the novelty of tion in the Pedersen et al. (2011) study and improved verbal parturition and newborns could produce aggression toward learning but not working memory (Feifel et al., 2012). Patients or avoidance of offspring. As is reviewed by MacDonald with schizophrenia have deficits in these cognitive domains (in this volume), OT anxiolysis is not restricted to reproduc- that profoundly impair their function and which do not tively relevant states in animals, has been demonstrated in respond to currently available antipsychotic medications human subjects, and may be therapeutically beneficial in (Harvey et al., 2006; Sergi et al., 2007; Penn et al., 2009). human anxiety disorders. Later in placental mammalian ArecentpreliminarytrialsuggeststhatOTmayhaveefficacy in reproductive, central OT systems evolved to facilitate the treating a quite different clinical disorder-addiction. Pedersen formation of selective maternal–infant bonds by enabling et al. (2013) discovered that twice daily intranasal OT for 3 days mothers to learn the unique odors of their offspring. was markedly more effective than placebo in reducing with- Chapter 2 also speculates that the evolution of avid and drawal symptoms and the amount of benzodiazepine required sustained maternal behavior, as well as lactation and in utero for successful medical detoxification of alcohol dependent fetal development, enabled, in multiple ways, the subsequent patients. This is the first evidence in human subjects that OT evolution of higher intelligence. The many recent reports that has therapeutic efficacy in a substance dependence disorder. intranasal OT administration facilitates social cognition in Remarkably, intranasal OT treatment produced no increased human subjects (see Evans et al. (in this volume) as well as side effects or adverse events in these clinical trials or other the review of these findings later in this chapter) suggest that studies (MacDonald et al., 2011). OT may have been centrally involved in advances in cognitive The sections below begin with reviews of animal and ability, especially social intelligence, during the evolution of human studies that led investigators to test OT as a treat- placental mammals. ment for schizophrenia and alcohol withdrawal. The clinical The primary goal of this article is to review evidence that trials are then described as well as relevant studies that have OT is a promising new treatment for schizophrenia and been published more recently. Questions are then posed alcohol dependence. A secondary but still key aim is to about the large gaps in our current knowledge about OT propose that the seemingly unrelated antipsychotic and systems in the human brain. These are discussed to highlight antiaddiction effects of OT may provide new insights into important areas for future research that will advance our other OT mechanisms that may have been selected during understanding of OT in schizophrenia, alcohol dependence placental mammalian evolution to facilitate the formation of and other addictive disorders as well as other types of social bonds. This argument is summarized here and restated psychopathology (see Macdonald and Feifel (2013) for a at the end of the article. As is discussed in more detail below, comprehensive overview of unresolved issues relevant to the antipsychotic efficacy of OT may be based on enhance- the development of OT pharmacotherapeutics). Finally, roles ment of sensory-motor gating, i.e., the ability to rapidly for which OT may have been selected during the evolution of process the significance of sensory input and initiate appro- placental mammalian maternal behavior are considered as priate behavioral responses. Successful evolution of avid and possible origins for OT efficacy in treating disorders as diverse sustained maternal behavior in placental mammals may as schizophrenia and alcohol dependence. have required a considerable up-grade in sensory-motor gating ability. Effective mothering requires accurate percep- tion of diverse sensory cues from offspring and swift mobi- lization of appropriate caretaking behaviors. The antiaddic- 3. Oxytocin and schizophrenia tion effects of OT appear to be based on inhibition of tolerance formation to addictive substances. The evolution MacDonald and Feifel (2012) thoroughly reviewed evidence of the ability to form lasting social attachments may have for therapeutic effects of OT in schizophrenia. We will briefly required selection of brain mechanisms that prevent habitua- summarize and update that evidence and elaborate further tion to the highly rewarding stimuli from reproductively on OT as a treatment for social cognition and function deficits important conspecifics such as offspring or pair-bonded in schizophrenia. 104 brain research 1580 (2014) 102–123

3.1. The rationale for oxytocin treatment of psychotic drugs, clozaril and amperozide, but not haloperidol, were symptoms in schizophrenia reported to increase plasma OT concentrations in rats (Uvnäs-Moberg et al., 1992). In addition, OT interactions have The strongest evidence that OT may have antipsychotic been reported with neurotransmitter systems that have been efficacy was studies employing several animal models of linked to schizophrenia; most importantly dopamine psychosis. The first animal model in which OT was tested is (Baskerville and Douglas, 2010; Shahrokh et al., 2010; Succu -induced hyperactivity that simulates the excessive et al., 2007), glutamate (Hrabovszky and Liposits, 2008; Ninan, dopaminergic activity which has long been hypothesized in 2011), and serotonin (Emiliano et al., 2007). Estrogen has schizophrenia (Carlsson, 1977; Carlsson and Carlsson, 2006). strong effects on CNS OT and OT receptors (Jirikowski et al., OT administration diminished cocaine or methampheta- 1990; Bale et al., 1995; Shughrue et al., 2002; Patisaul et al., mine-induced hyperactivity (Sarnyai et al., 1990; Qi et al., 2003; Choleris et al., 2008) which may be relevant to sex 2008), effects that were localized to the striatum and nucleus differences in the onset and severity of schizophrenia. accumbens, brain areas that have been implicated in schizo- All but one published studies of OT parameters in the phrenia (Meyer-Lindenberg, 2010). Subsequent studies were human brain have found differences between patients with conducted in a rat model of psychosis in which prepulse schizophrenia and normal subjects. Two groups reported that inhibition (PPI) is disrupted by administering psychotomi- CSF concentrations of OT or OT-associated neurophysin were metic drugs such as amphetamine or apomorphine that elevated in patients (Beckmann et al., 1985; Linkowski et al., increase dopamine receptor stimulation or NMDA receptor 1984) but a third group found no difference in OT concentra- antagonists like phencyclidine or MK801 (Geyer et al., 2001; tions (Glovinsky et al., 1994). Morphometric differences in OT- Geyer and Ellenbroek, 2003). PPI is the capacity to suppress associated neurophysin-immunostaining cells as well as startle response to a strong stimulus (usually acoustic) when reduced cell density were also identified in the PVN of “warned” by a preceding mild stimulus (Braff et al., 2001; schizophrenic patient brains (Mai et al., 1993; Bernstein Geyer et al., 2001). PPI is mediated by neural circuitry that et al., 1998). Serum concentrations of OT-associated neuro- includes the nucleus accumbens, the hippocampus and the physin were significantly higher in patients, especially those basolateral amygdala, all brain areas that have been impli- with paranoia (Legros et al., 1992). Plasma OT concentrations cated in the pathophysiology of schizophrenia (Koch and were lower in polydipsic, hyponatremic schizophrenic Schnitzler, 1997, Swerdlow et al., 2001). PPI performance is patients compared with polydipsic, normonatremic patients, impaired in patients with schizophrenia which is thought to non-polydipsic, normonatremic patients or healthy controls indicate aberrant sensory-motor gating in this disorder (Braff and did not differ among the latter 3 groups (Goldman et al., et al., 2001). PPI deficits are decreased by antipsychotic 2008). Rubin et al. (2010) found no difference in plasma OT medication in psychotomimetic drug-treated rodents and in concentrations between women or men with schizophrenia patients with schizophrenia. Both exogenous and endogen- compared with normal controls but higher concentrations in ous OT have been shown to diminish PPI deficits in this female patients correlated significantly with lower total, model of psychosis, especially in animals given NMDA positive and general Positive and Negative Symptom Scale antagonists (Feifel and Reza, 1999; Caldwell et al., 2009). OT (PANSS; Kay et al., 1987) scores and nearly significantly has also been tested in a third animal model that is based on (p .06) with negative PANSS scores. In addition, an OT gene ¼ evidence that the NMDA receptor is involved in psychosis. variant (rs2740204) has been significantly associated with Specifically, chronic treatment with NMDA antagonists clinical improvement with clozapine (Souza et al., 2010). (PCP, ketamine, MK-801) produces a loss of social interest In the same study, some OT receptor gene polymorphisms (a prominent negative symptom) (Sams-Dodd, 1999; Qiao were associated with improvement of positive psychotic et al., 2001) and PPI deficits. Chronic PCP administration symptoms with clozapine treatment while other receptor gene was noted by Lee et al. (2005) to decrease anterior hypotha- polymorphisms were associatedwithsymptomseverity.It lamic OT mRNA and to increase OT receptor binding in the should also be noted that studies conducted in the Soviet Union central nucleus of the amygdala (CeA). Infusion of OT into the 3–4decadespriortotherecenteraofintranasalOTclinical CeA fully restored social interest in chronic PCP-treated rats. trials in schizophrenia found that repeated administration of The greater efficacy of OT in reversing NMDA antagonist- intranasal or intravenous OT doses to severely, but rather induced PPI deficits and chronic PCP-induced psychotic-like imprecisely described, mentally ill patients, some probably with behavioral deficits in rats suggests that OT may exert its schizophrenia, improved symptoms and prevented hospitaliza- antipsychotic effects through interactions with glutamatergic tions (Bujanow, 1972; Bakharev et al., 1986). systems. Koenig et al. (2005) found that offspring of rat mothers exposed to variable, unpredictable stress during 3.2. The rationale for oxytocin treatment of social the third trimester also had deficits in social interest and cognition deficits in schizophrenia PPI. As in the chronic PCP model, these animals had dimin- ished OT mRNA in the hypothalamic paraventricular nucleus Social dysfunction is among the most disabling consequence (PVN), the site containing the largest number of centrally- of schizophrenia (Pinkham et al., 2003; Yager and Ehmann, projecting OT neurons, and elevated OT receptor binding in 2006) and responds poorly to currently available antipsycho- the CeA (Lee et al., 2007). Bilateral infusion of OT into the CeA tic medications (Buchanan et al., 1998; Kirkpatrick et al., 2000; restored normal social interest levels in these animals. Bellack et al., 2004; Penn et al., 2009). The loss of brain volume Other findings in animals have suggested that OT may and decline in general mental abilities that occur in exert effects in schizophrenia. The atypical antipsychotic many individuals with schizophrenia appear to contribute brain research 1580 (2014) 102–123 105

substantially to social dysfunction (Cahn et al., 2006). How- the trustworthiness of faces and empathy (Domes et al., ever, recent evidence indicates that aspects of mentation that 2007a; Fischer-Shofty et al., 2010; Guastella et al., 2010; facilitate social decisions and behavior are particularly Hurlemann et al., 2010; Petrovic et al., 2008; Theodoridou impaired in schizophrenia (Phillips et al., 2003; Pinkham et al., 2009). Plasma OT concentrations were reported to et al., 2003; Yager and Ehmann, 2006; Penn et al., 2008). This increase in participants in a trust game that received a high type of cognition, “social cognition,” is comprised of emotion monetary transfer indicating trust and who then reciprocated recognition (e.g., identifying the emotional states of others generously (Zak et al., 2005). During trust-related interactions, from their facial expressions and other social cues), attribu- plasma OT levels rose in healthy controls but not in patients tional style (beliefs about the causes of events) and theory of with schizophrenia (Keri et al., 2009). In schizophrenia mind (inferring the thoughts and feelings of others), areas patients, plasma OT levels correlated positively with their that are consistently impaired in schizophrenia (Green et al., accuracy in identifying facial emotions (Goldman et al., 2008). 2005; Penn et al., 2006; Bora et al., 2009; Kohler et al., 2010), For more comprehensive reviews and critique of studies of and which are related to real-world outcomes, such as social intranasal OT on human social behavior and social cognition, and community functioning(Couture et al., 2006; Pinkham including numerous studies since 2010, see Guastella and and Penn, 2006; Fett et al., 2011). Unfortunately, there is no MacLeod (2012) and Evans et al. (in this volume). evidence that psychopharmacological treatment, such as Confidence that OT administered by the intranasal route currently available antipsychotic medications, improves enters the brain was increased by results published by Born social cognition in schizophrenia. Specifically, three large et al. (2002). They found that CSF levels of vasopressin were clinical trials (including the CATIE trial) found that typical markedly increased 10 min after intranasal administration and atypical antipsychotic medications did not improve and remained elevated for at least 80 min. This has led to the social cognition in schizophrenia (Harvey et al., 2006; Sergi assumption that OT, which is structurally very similar to et al., 2007; Penn et al., 2009). vasopressin, also readily enters the CNS after intranasal Decades of research in animals have established that CNS administration. Several fMRI studies have found that activa- OT has many effects on social behavior including activation tion of the amygdala in subjects viewing unpleasant or of key social attachments such as maternal behavior in frightening images is altered by intranasal administration of several species, pair-bond formation in females of some OT (Kirsch et al., 2005; Domes et al., 2007b, 2010; Zink monogamous species and olfactory-based social memory, a and Meyer-Lindenberg, 2012). These results strongly support rodent model of social cognition (Pedersen et al., 1992; Carter the conclusion that OT enters the CNS after intranasal et al., 1999; Gimpl and Fahrenholz, 2001; Lee et al., 2009). Over administration. the past decade, human studies in which OT was adminis- tered in a nasal spray have found a wide range of pro-social 3.3. Clinical trials of oxytocin treatment of schizophrenia effects (MacDonald and MacDonald, 2010). The advent of these studies was probably delayed by the inconsistent All 3 published studies were randomized, double-blind and results of early studies employing the intranasal route of placebo-controlled trials conducted in patients who met administration to determine if OT had amnestic effects criteria for DSM-IV schizophrenia, were at least moderately previously found in animals (Ferrier et al., 1980; Fehm- symptomatic (total PANSS scores Z55) and had been clini- Wolfsdorf and Born, 1991; Bruins et al., 1992) and by the cally stable and on the same doses of antipsychotic medica- negative results of two clinical trials testing whether sus- tions (atypicals in almost all cases) for at least 4 weeks prior tained daily intranasal OT treatment improved symptoms in to study entry. These studies found no differences between obsessive-compulsive disorder (den Boer and Westenberg, OT and placebo groups in side effects (including extrapyr- 1992; Epperson et al., 1996). amidal symptoms) or laboratory measures over the treatment The current era of studying OT effects on human social periods. behavior was ushered in by a particularly elegant and highly Feifel et al. (2010) were the first to conduct a clinical trial of influential experiment conducted by Kosfeld et al. (2005) intranasal OT treatment in patients who met DSM-IV criteria which demonstrated that a single intranasal dose of OT for schizophrenia. The study had a cross-over design and significantly increased interpersonal trust. Numerous studies enrolled 15 adult patients (12 male, 3 female). The order of published in the ensuing 5 years reporting that intranasal OT intranasal treatments (OT vs. placebo) was randomly produced a variety of social behavioral and social cognitive assigned during an initial 3-week treatment period after effects in human subjects. These included mostly positive which subjects received the other treatment during a sub- influences on social behavior including replication of the sequent 3-week period following a 1-week washout. Intrana- Kosfeld et al. (2005) finding of interpersonal trust enhance- sal OT (Syntocinon Spray, Novartis, Basel, Switzerland) was ment (Baumgartner et al., 2008; Mïkolajczak et al., 2010a, dosed at 20 IU (5 insufflations) twice daily (BID) for 1 week 2010b), increase in eye contact (Andari et al., 2010; Guastella and then 40 IU BID for 2 weeks. PANSS total scores, PANSS et al. 2008), improved perception of and responsiveness to positive subscale scores, PANSS negative subscale scores as social cooperation in patients with autism spectrum disorder well Clinical Global Impressions-Improvement (Guy and (Andari et al., 2010) but also amplification of envy and Bonato, 1970) scores were significantly lower in the OT gloating at others' misfortune (Shamay-Tsoory et al., 2009). compared to the placebo group at treatment week 3 but not Other intranasal OT studies published during this time period at 1 or 2 weeks (see Figs. 1–3). PANSS general subscale scores found improvements in various domains of social cognition: did not differ between treatment groups at 3 weeks (Fig. 4). face emotion recognition; theory of mind; and perception of There was a significant drug by time interaction for PANSS 106 brain research 1580 (2014) 102–123

total scores indicating there was greater decline in this Feifel et al. (2010) measure from baseline to 3 weeks in the OT compared to 85 placebo group. + Neurocognitive deficits, including impaired verbal and 80 working memory, are common in schizophrenia and respond poorly to current antipsychotic medications (Keefe and 75 Harvey, 2012). Intranasal administration of OT has been Oxytocin reported to have amnestic effects in some but not all human

PANSS TOTAL PANSS Placebo 70 studies (Fehm-Wolfsdorf and Born, 1991; Bruins et al., 1992). Therefore, Feifel et al. (2012) also administered the California 65 Verbal Learning Test (CVLT, Delis et al., 2000) and the Letter Base 1 2 3 Number Sequencing Task (LNS, Wechsler, 1997), a test of Weeks working memory, in the cohort of 15 described above at baseline and at week-3 of each test treatment period. During Pedersen et al. (2011) administration of the CVLT, subjects read a list of 16 words 85 over 5 trials, and then engage in recall tasks, including free recall, recall after an interference list is read, recall after 80 different amounts of elapsed time (i.e. short and long delay) a and discriminating between words from different lists. Each 75 task generates a separate subscore. Administration of the Oxytocin

LNS Task involves reading to subjects a series of lists of TOTAL PANSS 70 Placebo letters and numbers in mixed orders ranging in length from 2 # fi to 8. After each list is read, subjects are asked to recall rst 65 the digits and then the letters in the order they were Base 2 presented. Scores are the total number of correct responses Weeks on all of the lists. The CVLT total recall, short delay free recall and total recall discriminability measures were significantly Modabbernia et al. (2013) 95 improved by OT compared to placebo treatment and com- pared to baseline, although most of the improvement a b occurred in subjects who received OT during the second 90 b treatment period (Fig. 5). OT did not improve performance on the LNS Task compared to placebo or baseline (Fig. 5). The 85 effects of 3 weeks of intranasal OT treatment on CVLT found Oxytocin by Feifel et al. (2012) in patients with schizophrenia contrasts

PANSS TOTAL PANSS Placebo with previous reports that a single intranasal dose of OT 80 worsened verbal memory (Ferrier et al., 1980; Fehm-Wolfsdorf et al., 1984; Bruins et al., 1992; Heinrichs et al., 2004). However, 75 these results suggest that OT treatment has the potential to Base 2468 improve some neurocognitive deficits in schizophrenia. Weeks Pedersen et al. (2011) reported the results of a 2-week Fig. 1 – Mean PANSS total scores at baseline and at each treatment trial conducted in 20 subjects (17 male and 3 assessment time point for the intranasal oxytocin and placebo female) with paranoid or undifferentiated schizophrenia treatment groups in three randomized clinical trials in patients who received BID intranasal doses of OT (24 IU/dose, N 11) ¼ with schizophrenia. a Denotes po.05, b denotes po.001 or placebo (N 9). The PANSS, the Paranoia Scale (Fenigstein ¼ between treatment groups change from baseline, denotes and Vanable, 1992), the Brüne Theory of Mind Picture Stories þ po.001 between treatment groups at 3 weeks, and # denotes Task (Brüne, 2003) and the Trustworthiness Task (Adolphs p .0009 within treatment group change from baseline. et al., 1998) were administered at baseline and after 14 days of ¼ test treatments. Five additional subjects subsequently com- pleted the protocol increasing the total cohort to 25 (OT 14 ¼ and placebo 11). Data analyzed in the larger cohort also (Fig. 3) and general psychopathology (Fig. 4) subscale scores ¼ included mean scores on 5 PANSS items relevant to social fell significantly within the OT treatment group but not function (the average of the suspiciousness, hostility, pas- within the placebo group. Similarly, significant reductions in sive/apathetic social withdrawal, uncooperativeness and PANSS anxiety item scores and Paranoia Scale scores active social avoidance items). The results from the larger occurred solely in OT recipients (po.05 for both measures). sample are presented here and cited as Pedersen et al. (2011). However, PANSS suspiciousness item scores declined signifi- The OT compared to the placebo treatment group had a cantly in both treatment groups. Also restricted to OT significantly greater decline in total PANSS scores from base- recipients were significant declines in PANSS social items line to 2 weeks (Fig. 1) as well as a trend toward a significant mean scores (Fig. 6). decline in PANSS general subscale scores (p .09). PANSS total The Brüne Theory of Mind Picture Stories Task (Brüne, ¼ scores (Fig. 1) as well as PANSS positive (Fig. 2), negative 2003) includes six 4-panel cartoons depicting interpersonal brain research 1580 (2014) 102–123 107

Feifel et al. (2010) Feifel et al. (2010) 24 22 23 + + 21 22 ++ 20 21

19 20 Oxytocin Oxytocin Placebo 19 PANSS Negative PANSS Positive PANSS Placebo 18 18 17 17 Base 1 2 3 Base 1 2 3 Weeks Weeks Pedersen et al. (2011) Pedersen et al. (2011) 21 23

22 20

21 19 # 20 18 19 Oxytocin

Oxytocin PANSS Negative 18 17 Placebo PANSS Positive PANSS Placebo 17 # 16 Base 2 16 Base 2 Weeks Weeks Modabbernia et al. (2013) Modabbernia et al. (2013) 26 25

24 b b 25 23

22 b b 24 21 Oxytocin PANSS Negative Oxytocin Placebo 20

PANSS Positive PANSS Placebo 19 23 Base 2 4 6 8 18 Weeks Base 2 4 6 8 Weeks Fig. 3 – Mean PANSS negative subscores at baseline and at each assessment time point for the intranasal oxytocin and Fig. 2 – Mean PANSS positive subscores at baseline and at placebo treatment groups in three randomized clinical trials each assessment time point for the intranasal oxytocin and in patients with schizophrenia. b Denotes po.001 between placebo treatment groups in three randomized clinical trials treatment groups change from baseline; denotes p .023 in patients with schizophrenia. b Denotes po.001 between þ ¼ between treatment groups at 3 weeks; and # denotes p .025 treatment groups change from baseline; denotes p .006 ¼ þ ¼ within treatment group change from baseline. between treatment groups at 3 weeks, and # denotes p .003 ¼ within treatment group change from baseline. Task (Adolphs et al., 1998) involves rating each face depicted in a series of pictures for trustworthiness on a scale ranging interactions that are sometimes deceptive and often produce from 3 to 3. In the study conducted by the Pedersen group, À þ false beliefs in specific characters. Questions are asked to Trustworthiness Task scores were based on ratings of determine subjects' understanding of the beliefs of cartoon untrustworthy faces (below the mean in a normal sample). characters and their intentions. As illustrated in Fig. 6, sub- The OT treatment group had a decline in untrustworthiness jects in the OT group, but not subjects in the placebo group, ratings that trended toward significance while untrustworthi- improved significantly in accurately identifying second order ness ratings changed very little in the placebo group (Fig. 6). false belief (i.e., cartoon characters understood that other Modabbernia et al. (2013) conducted an 8-week clinical characters had false beliefs) and in accurately identifying trial in which 40 subjects (33 male and 7 female) with third order false belief (i.e., index cartoon characters under- paranoid, residual or undifferentiated schizophrenia, all of stood that other characters erroneously believed the index whom were on stable doses of risperdone, were randomized characters had false beliefs). OT recipients also exhibited a to OT (N 20) or placebo (N 20); 37 subjects completed the ¼ ¼ trend toward significant improvement in accurately detecting entire protocol. OT recipients received 20 IU BID during the deception in the Brüne Task (Fig. 6). The Trustworthiness first week and 40 IU BID for the remaining 7 weeks. PANSS 108 brain research 1580 (2014) 102–123

Feifel et al. (2010) et al. (2010) and the Pedersen et al. (2011) trials but somewhat 41 higher in the Modabbernia et al. (2013) trial. Perhaps the most notable discrepancy among these trials is that OT treatment 39 produced the greatest decline from baseline in total PANSS scores (9.8, 12.4%) in the Pedersen et al. (2011) study that administered the lowest dose of OT (24 IU BID) for the short- 37 est period (2 weeks). Oxytocin treatment for 3 weeks and at a higher dose for the last 2 weeks (40 IU BID) resulted in a

PANSS General 35 smaller reduction in total PANSS scores (7.9, 9.7%) in the Oxytocin Feifel et al. (2010) study, while OT treatment for 8 weeks and Placebo 33 at a higher dose for the last 7 weeks (40 IU BID) produced a yet Base 1 2 3 smaller decline in total PANSS scores (6.7, 7.5% [this percen- Weeks tage is based on data in the report but is lower than the 11.2% Pedersen et al. (2011) claimed by the authors]) in the Modabbernia et al. (2013) 41 study. In contrast to the Pedersen et al. (2011) trial, 2 weeks of OT treatment did not significantly decrease PANSS scores in

39 either the Feifel et al. (2010) or Modabbernia et al. (2013) studies. While OT produced significant declines in PANSS scores in all 3 trials, the discrepancies in the magnitude and 37 Oxytocin latencies of the effects on psychotic symptoms underscore Placebo the importance of directly comparing a range of doses in

PANSS General 35 future studies. The improvements obtained by Pedersen et al. # (2011) and Feifel et al. (2012), respectively, in some social 33 cognition and neurocognition measures are very exciting Base 2 because deficits in these areas contribute to a high rate of Weeks disabling social dysfunction in schizophrenia and are not Modabbernia et al. (2013) improved by currently available antipsychotic medications. 44 Longer treatment trials will be necessary to determine if OT treatment produces further improvements in social cognition and neurocognition and perhaps reduces social dysfunction. 42 a a See MacDonald and Feifel (2012) for an in-depth discussion of other important issues about the therapeutic use of OT in 40 schizophrenia. Oxytocin Placebo

PANSS General 38 3.4. Other recent oxytocin studies in schizophrenia

fi 36 There have been additional publications since the rst clin- Base 2 4 6 8 ical trial report by Feifel et al. (2010) that for the most part Weeks further support a significance role for OT in the pathophy- Fig. 4 – Mean PANSS general subscores at baseline and at siology and pharmacotherapy of schizophrenia. In a compar- each assessment time point for the intranasal oxytocin and ison of intranasal administration of two doses of OT (10 or placebo treatment groups in three randomized clinical trials 20 IU) or placebo among 5 polydipsic schizophrenia patients, in patients with schizophrenia. a Denotes po.05 between 8 non-polydipsic patients and 11 healthy controls, Goldman treatment groups change from baseline; and # denotes et al. (2011) found that the lower OT dose decreased accurate fi p .011 within treatment group change from baseline. identi cation of face emotions in patients while the higher ¼ dose increased accurate emotion identification in polydipsic patients, an effect largely attributable to decreased bias to scores were obtained at baseline and every 2 weeks during the identify fear in non-fearful faces. Averbeck et al. (2012) treatment period. As in the Feifel et al. (2010) and Pedersen et al. reported that a single intranasal dose of OT (24 IU) compared (2011) studies, PANSS total and subscale scores did not differ to placebo in a cross-over design significantly improved between treatment groups at baseline. The OT compared to the accurate emotion recognition in patients with schizophrenia placebo group had significantly greater declines from baseline shown images of morphed and unmorphed faces showing a in (1) PANSS total scores at 4, 6 and 8 weeks of treatment (Fig. 1); range of emotions (anger, disgust, fear, happiness, sadness, and (2) PANSS positive (Fig. 2), negative (Fig. 3)andgeneral and surprise) although OT did not significantly increase subscale (Fig. 4)scoresat6and8weeks. accuracy of identifying specific emotions. The results of these 3 trials support the conclusion that Lee et al. (2013) conducted a placebo-controlled clinical intranasal administration of OT decreases psychotic symp- trial in 28 subjects (12 inpatient and 16 outpatient) with toms in schizophrenia. However, this conclusion must be schizophrenia or schizoaffective disorder testing BID intra- considered preliminary because of the small size of the nasal OT (20 IU/dose) for 3 weeks on general psychiatric studies. Baseline PANSS scores were similar in the Feifel symptoms including a 4-item positive symptom score on brain research 1580 (2014) 102–123 109

Total Recall, Trials 1-5 Total Recall Discriminability 2.5 60 * ## * ##

55

50 2

45

40 1.5

0 0 Baseline Oxytocin Placebo Baseline Oxytocin Placebo

Short Delay Free Recall Long Delay Free Recall 12 * ## 12 # #

10 10

8 8

6 6

0 0 Baseline Oxytocin Placebo Baseline Oxytocin Placebo

Total Recognition Discriminability 10 Letter Number Scale 3.5

3

8

2.5

2 6 0 0 Baseline Oxytocin Placebo Baseline Oxytocin Placebo

Fig. 5 – Mean ( SEM) performance on various measures on the California Verbal Learning Test (A–E) and the Letter Number þ Sequencing Task (F) at baseline and after 3 weeks of intranasal oxytocin or placebo treatment. Figure obtained from Feifel et al. (2012) with permission. * Denotes po.05 difference from placebo; # denotes po.05, and ## denotes pr.001 difference from baseline. the Brief Psychiatric Rating Scale (BPRS, Overall and Gorham, obtained at baseline as well as at 1 and 3 weeks during the 1962), negative symptom ratings on the modified Scale for the test treatment period. Total BPRS scores decreased signifi- Assessment of Negative Symptoms (SANS, Buchanan et al., cantly more in the placebo than the OT group although there 2007) as well as olfactory identification ability using the was no difference between treatment groups on change in University of Pennsylvania Smell Identification Test (UPSIT, positive symptoms. On the SANS, total scores and motiva- Doty et al., 1984). Olfactory identification ability has been tion-pleasure subscores were significantly better in inpatients reported to be particularly impaired in patients with negative after 3 weeks of treatment with OT compared to placebo. symptoms and the deficit syndrome (Corcoran et al., 2005; However, these measures did not differ between treatment Malaspina and Coleman, 2003; Moberg and Turetsky, 2003; groups at 3 weeks in the combined inpatient and outpatient Moberg et al., 2006; Strauss et al., 2010). Measurements were sample. Among all subjects, total UPSIT scores and pleasant 110 brain research 1580 (2014) 102–123

b b 3 3

2 2

1 1

0 0 Base 2 weeks Base 2 weeks Base 2 weeks Base 2 weeks Oxytocin Placebo Oxytocin Placebo

Trustworthiness Task 0

a -5 3 -10

2 -15

-20 1 -25 a

0 -30 Base 2 weeks Base 2 weeks Base 2 weeks Base 2 weeks Oxytocin Placebo Oxytocin Placebo

4 PANSS Social Items

3 c

2

1

0 Base 2 weeks Base 2 weeks Oxytocin Placebo

Fig. 6 – Mean ( SD) performance on various measures on the Brüne Theory of Mind Picture Stories Task (A–C) and on the þ Trustworthiness Task (D) as well as the mean ( SD) PANSS social item scores (E) at baseline and after 2 weeks of intranasal þ oxytocin or placebo treatment. a Denotes po.08, b denotes po.05, and c denotes po.001 difference from baseline.

smell subscores improved significantly in the OT compared to healthy controls matched for age and gender, 2 single the placebo group; there were no treatment group differences nucleotide polymorphisms (SNPs) of the OT receptor gene, on neutral or unpleasant smell subscores. rs53576 and rs237885, were found to be significantly asso- Higher plasma OT concentrations were related to perceiv- ciated with a diagnosis of schizophrenia and, respectively, ing images of faces as happier in women with schizophrenia with PANSS general and negative subscores in the patients and healthy controls but not in male patients or controls (Montag et al., 2013). The same group also found that (Rubin et al., 2011). Sasayama et al. (2012) found that CSF another SNP of the OT receptor gene, rs2254298, was asso- concentrations of OT in male schizophrenia patients corre- ciated with “empathic concern” but not with other dimen- lated negatively with doses of atypical but not typical sions of empathy measured on the Interpersonal Reactivity antipsychotic medications and also correlated negatively Index (Davis, 1983) in matched cohorts of schizophrenia with PANSS negative subscale scores even after controlling patients and controls and, in the patients, was related to for atypical antipsychotic doses. In a case-control design, PANSS general psychopathology subscale scores (Montag comparison of 406 patients with schizophrenia and 406 et al., 2012). brain research 1580 (2014) 102–123 111

3.5. Questions to guide future research on oxytocin antisera have not been developed (Yoshida et al., 2009). This mechanisms in schizophrenia has been a major obstacle to identifying OT-activated neural circuits involved in animal social behavior and will similarly MacDonald and Feifel (2012) have summarized important impede investigation of OT mechanisms in preclinical animal areas for future research concerning OT pharmacotherapy models of psychosis, addiction and other psychopathologies. in schizophrenia. Below we discuss other questions that are These methodological barriers to elucidate OT receptor loca- fundamental for understanding OT involvement in schizo- lization and neural circuits that mediate OT effects in the phrenia and other psychopathology. CNS are discussed in detail by Pedersen (2013). Furthermore, attempts to develop PET ligands to locate OT receptors in the 3.5.1. Where are oxytocin receptors in the human brain? brains of living human subjects have had disappointing It is of great importance to identify with precision where OT results (Smith et al., 2012). receptors are expressed in the human brain, especially in In a promising new development, Boccia et al. (2013) have areas that have been implicated in schizophrenia and other conducted the first study clearly demonstrating and locating psychopathology. In reviews of OT in human behavior and OT receptors in the human brain with immunohisto- psychopathology, assertions are routinely made about the chemistry employing a monoclonal antibody developed by locations of central OT receptors. These assertions are based Takemura et al. (1994) that selectively binds OT receptor in solely on autoradiographic studies in rodents and, to some human reproductive tissue. At autopsy, blocks of fixed brain extent, PCR studies in sheep or on early autoradiography tissue containing areas of interest were dissected from two studies in human brain that employed a radioligand that has deceased human females (44 and 28 years old). OT receptor since been shown to have poor selectivity for the human OT immunostaining was visualized in discrete cell bodies and/or receptor (Loup et al., 1989, 1991; Toloczko et al., 1997). There fibers in the central and basolateral regions of the amygdala are no definitive, replicated studies of OT receptor locations (see front cover of this issue), medial preoptic area, anterior in the human brain or mammalian species other than and ventromedial hypothalamus, olfactory nucleus, vertical rodents (rat, mouse, and vole) and sheep. Among rodent limb of the diagonal band, ventrolateral septum, anterior and other mammalian species, central OT receptor distribu- cingulate cortex and hypoglossal and solitary nuclei. OT tion varies widely (Tribollet et al., 1992; Hammock and Young, receptor staining was not observed in the hippocampus 2006). Therefore, findings from rodents and sheep can, at (including the CA2 and CA3), parietal cortex, raphe nucleus, best, suggest where OT receptors might be located in the nucleus ambiguus or pons. Unfortunately, tissue blocks that human brain. included regions that exhibit volumetric or connectivity Perhaps the best available evidence of where OT receptors abnormalities in schizophrenia, such as the prefrontal and may be in the human brain comes from fMRI studies of the superior temporal cortices as well as basal ganglia (Meyer- effects of intranasal OT administration on brain activation by Lindenberg, 2010; Fitzsimmons et al., 2013), were not col- social stimuli or social interactions. The brain sites where OT lected for this study. The Boccia et al. (2013) results suggest has altered activation have varied among studies possibly that there are some similarities, but also important differ- because of differences in test conditions (Meyer-Lindenberg ences, in the locations of OT receptors in human and rodent et al., 2011; Zink and Meyer-Lindenberg, 2012). The amygdala brains. is the only area where activation was affected by OT in all studies. Also, OT has had strikingly contrasting effects on 3.5.2. Does oxytocin regulate brain development? brain activation in men and women performing similar face Several lines of evidence suggest that OT is involved in brain emotion processing tasks (Domes et al., 2007b, 2010). OT development from early in ontogeny. The initial hints that decreased activation in the right amygdala and a few addi- this may be so were autoradiographic studies showing quite tional sites in men but increased activation of the left different OT binding patterns in the brains of neonatal amygdala as well as numerous other sites in women. Infer- compared to adult rat brains (Shapiro and Insel, 1989; ences about the location of OT receptors from these investi- Tribollet et al., 1989). Binding emerged early in the postnatal gations are limited to brain areas that are activated by social period in brain sites in which abnormalities have been found stimuli. Future studies are needed that examine OT effects on in schizophrenia. These include the cingulate cortex, caudate brain activation by a wider range of stimuli as well as putamen, dorsal hippocampus and thalamus. Binding in determining OT effects in the absence of stimuli. these sites reached peak density at approximately age 14 Unfortunately, selective radioligands for the human OT days and then diminished and was no longer discernible by receptor have not yet been developed. Therefore, autoradio- the end of the third postnatal week except for the persistence graphic location and quantification of OT binding have not of weaker binding in the caudate putamen. The adult pattern been possible in postmortem human brains. Even though of OT binding did not emerge until puberty. OT receptor antibodies that selectively bind human OT receptors have mRNA has been found in the rat telencephalon as early as been used for two decades to visualize receptors in human fetal day 12 although it is unclear if the receptor is expressed reproductive tissues, they have only recently been employed that early (Chen et al., 2000). Analogous with the results of to conduct immunohistochemistry localization of OT recep- autoradiography studies in postnatal rat brains, we have tors in the human brain (see below). Surprisingly, immuno- preliminary, unpublished evidence that OT receptor immu- histochemical visualization of central OT receptors in other nostaining in second trimester human fetal brain is located in species, including rodents, has yet to be consistently sites, such as the hippocampus, where immunostaining was achieved because reliable, selective OT receptor-directed not found by Boccia et al. (2013) in the adult human brain. 112 brain research 1580 (2014) 102–123

These findings suggest that OT may play a role in early brain 3.5.3. Does oxytocin ameliorate brain connectivity and development including brain areas that have been implicated plasticity deficits in schizophrenia? in schizophrenia. While there is considerable indirect evi- Many imaging studies have found impaired connectivity dence (see below), experiments to directly test this important between numerous brain areas in schizophrenia, especially hypothesis and to determine the specific neurodevelopmen- between prefrontal and limbic regions (Fitzsimmons et al., tal processes that may be regulated by OT have not yet 2013; Schmitt et al., 2011). Considerable evidence suggests been conducted in rodents or other species. Research in that neural plasticity, the ability of synapses to change their this area could provide important new insights into activity in response to altered afferent input, is also impaired mechanisms of great significance to the pathophysiology of in schizophrenia (Daskalakis et al., 2008; Goto et al., 2010; schizophrenia. Hasan et al., 2011; Stephan et al., 2006, 2009). Cavus et al. A developmental role for OT in the fetal brain is supported (2012) recently reported that a 2-min period of high frequency by the marked social behavior abnormalities exhibited by visual stimulation increased subsequent visual evoked mice with a null mutation of the OT gene (OT / , gestated potentials (VEPs) in healthy control subjects but failed to À À in OT / mothers) or mice with a null mutation of the OT increase VEPs in patients with schizophrenia. À À receptor gene, in which OT activity was absent from concep- Several fMRI studies have found that intranasal OT treat- tion, compared to wild type (WT) mice or OT( / ) mice ment has effects on connectivity between some brain areas À À gestated in OT ( / ) mothers that were exposed to OT in implicated in schizophrenia that vary with the test situation þ À utero (Takayanagi et al., 2005). Further comparison with WT (Bethlehem et al., 2013; Rilling et al., 2012; Sripada et al., 2013; mice has identified a number of deficits in male OT receptor Riem et al., 2012; Wittfoth-Schardt et al., 2012). Animal knockout (OTRKO) mice in reciprocal social interactions and studies have demonstrated that OT alters neural plasticity in social communication in adulthood (Pobbe et al., 2012a, in brain areas that exhibit abnormalities in schizophrenia 2012b). In agreement with the initial report by Takayanagi such as the prefrontal cortex (PFC) and hippocampus. Speci- et al. (2005), Dhakar et al. (2012) found elevated aggressive fically, OT perfusion of mouse brain slices for 1 hour con- behavior in male OTRKO mice but no aggression differences verted long-lasting depression into long-lasting potentiation from WT mice in conditional forebrain OTRKO mice in which of glutamatergic neurotransmission in the infralimbic, med- the gene was not excised until postweaning (ages 21–28 days). ial PFC which is thought to play an important role in Additional evidence has pointed toward a brain developmen- regulating social cognition (Ninan, 2011). Also, OT perfusion tal role of OT in the postnatal–prepubertal period. A number of hippocampal slices enabled long-lasting, long-term poten- of studies in monogamous prairie or Mandarin voles have tiation by one-train, rather than multiple-train, tetanus established that a single dose of OT or OT antagonist shortly stimulation (Tomizawa et al., 2003). fMRI could be used to after birth alters, in a species, sex and dose dependent determine if intranasal OT administration during a clinical manner, adult expression of various dimensions of social trial “normalizes” connectivity in the brains of schizophrenia behavior (Bales and Carter, 2003a, 2003b; Bales et al., 2004, subjects and if changes in connectivity are related to symp- 2007; Jia et al., 2008a, 2008b). OT regulation of human brain tom improvement. The elegant method of Cavus et al. (2012) development is suggested by reports of relationships of could be employed to determine if intranasal OT treatment specific OT gene or OT receptor gene SNPs with autism, improves neural plasticity in patients. symptoms in schizophrenia, profound abnormalities of child social behavior as well as emotionality, social behavior and 3.5.4. Is OT regulation of stress relevant to the onset of stress responses (Beitchman et al., 2012; Wu et al., 2005; schizophrenia? Jacob et al., 2007; Liu et al., 2010; Souza et al., 2010; Ebstein Abundant evidence indicates that stress contributes to the et al., 2012; Malik et al., 2012). onset of psychosis in adolescents and young adults at high If OT regulates early brain development, deficits in OT or genetic risk for developing schizophrenia (Walker and Diforio, OT receptor expression could contribute to neurodevelop- 1997; Walker et al., 2008; McEwen, 2000; Broome et al., 2005; mental abnormalities that contribute to the onset of schizo- van Os et al., 2005, 2010; Thompson et al., 2007a, 2013b; phrenia. A plethora of evidence has linked a wide variety of Arnsten, 2009; Brown, 2011). Stress responses increase during environmental insults from conception to adolescence to the normal postpubertal period. In agreement with findings increased risk of developing schizophrenia (van Os et al., in animals (Goldman et al., 1973; Vazquez and Akil, 1993; 2010; Brown, 2011). Brown (2011) has concluded that a large McCormick and Mathews, 2010), adolescents exhibit exagger- portion, if not the majority of schizophrenia cases can be ated stress responses compared to adults including (1) accounted for by interactions between environmental and increased activity of the hypothalamic-pituitary-adrenal genetic factors and by other mechanisms involving the subtle (HPA) axis around puberty (Walker et al., 2004), with a marked interplay between environments and genes. Many of these rise in salivary and urinary cortisol around 13 years of age environmental insults occur during pregnancy or early child- that continues throughout adolescence (Kenny et al., 1966), hood (infections, toxins, nutritional deficits, etc.). Of great and (2) greater social stress increases in heart rate and importance will be determining whether epigenetic mechan- cortisol release (Buske-Kirschbaum et al., 1997). Increased isms, such as methylation, impair OT or OT receptor gene stress-reactivity in adolescence has been attributed to imma- expression in schizophrenia and if the degree of impairment turity of fronto-limbic circuits. Changes from shorter to is related to exposure to environmental risk factors. A high longer-range connections that increase connectivity between degree of methylation of the OT receptor gene has been limbic and prefrontal regions and enable greater efficiency in reported in autism (Gregory et al., 2009). top-down regulation of limbic activity occur during this brain research 1580 (2014) 102–123 113

developmental period (Fair et al., 2007; Kelly et al., 2008). cytokines from those cells (Szeto et al., 2008). IV infusion of Higher-order prefrontal areas involved in emotion and stress OT potently blocks endotoxin-induced release of a wide range regulation mature relatively late during adolescence, contri- of pro-inflammatory cytokines and other immune-activating buting to emotion dysregulation and increased sensitivity to factors (Clodi et al., 2008). stress and emotion during this period (Galvan et al., 2006; Also relevant are animal studies finding that OT regulates Ganzel et al., 2007; Casey et al., 2010). mesolimbic dopamine activity through effects in the ventral Dysregulated stress activation of the HPA has also been tegmental area and the nucleus accumbens (Succu et al., reported during the initial transition to psychosis and in 2007; Melis et al., 2007; Shahrokh et al., 2010; Liu and Wang, asymptomatic individuals at high genetic risk for the disorder 2003). In agreement with these findings, a recent fMRI study (Walker and Diforio, 1997; van Os et al., 2005; Thompson found that intranasal OT administration increased VTA et al., 2007a, 2007b; Walker et al., 2008). During adolescence, activity in response to socially relevant visual cues (Groppe glucocorticoids exert programming effects on neurocircuitry et al., 2013). involved in higher cognitive functions. Sustained high levels These considerations suggest that our understanding of of glucocorticoids may result in enduring changes in adult the significance of stress for the onset of psychosis may be cognitive function (McCormick and Mathews, 2010). Conse- advanced by comparing endocrine, autonomic and brain quences of frequent or sustained activation of the HPA axis activation responses to stress during adolescence between also include damage to regions rich in glucocorticoid recep- individuals at high and low risk of developing schizophrenia. tors, such as the hippocampus (Hoschl and Hajek, 2001; Furthermore, investigating the effects of intranasal OT on McEwen, 2002); damage to prefrontal cortex and its function stress responses from early to late adolescence may shed (Arnsten, 2009); and poor hippocampal-frontal connectivity. light on OT involvement in the development of stress In addition to increasing glucocorticoid levels, acute stress responses during this critical period and determine whether exposure increases dopamine synthesis and release in the OT “normalizes” aberrant stress responses that may be found brain (Moghaddam, 2002; Czyrak et al., 2003; Dallman et al., in high risk individuals. The latter findings would suggest 2004 ), especially in the mesolimbic system (Finlay and that OT treatment of high risk individuals during adoles- Zigmond, 1997; Marinelli et al., 2006). Hippocampal and cence/early adulthood, especially those exhibiting prodromal prefrontal cortex pathophysiology leads to increased stress- subsyndromal psychotic symptoms, may protect them from induced dopamine release, which is associated with psycho- conversion to psychosis. tic symptoms and increased salience of, and greater reaction to, emotional stimuli (Broome et al., 2005; Grace, 2004). Exposure to psychosocial stress can compromise fronto- 4. The rationale for oxytocin treatment limbic circuits through excessive dopamine activity, leading of alcohol withdrawal to relatively increased limbic activity with concomitant dis- ruptions in prefrontal activity. Impairment in stress- Numerous studies in rodents found that OT treatment prior regulation neural pathways may synergize with increased to each successive dose of alcohol, opioids or cocaine inhib- sensitization to dopamine during adolescence (Jezierski et al., ited tolerance formation, i.e., the typical decline in efficacy of 2007) to further adversely impact cognition and neural the addictive substance following repeated administration of circuitry in high risk individuals. the same dose (Kovács et al., 1998). Specifically, peripheral or Animal studies have shown that OT decreases all compo- central administration of OT in mice blocked the develop- nents of the stress response: HPA axis, sympathetic nervous ment of tolerance to the hypothermic, myorelaxant, akinetic system and immune activation as well as anxiety (McCarthy and effects of alcohol (Rigter et al., 1980; Pucilowski et al., 1996; Windle et al., 1997; Neumann et al., 2000; Bale et al., 1985; Szabó et al., 1985, 1989; Jodogne et al., 1991 ). et al., 2001; Mantella et al., 2003, 2004; Amico et al., 2004; OTKO compared to WT mice exhibited significantly more Szczepanska-Sadowska, 2008; Wsol et al., 2008; Grippo et al., rapid tolerance formation to effects of repeated doses of 2009; Cohen et al., 2010; Jankowski et al., 2010). Further- alcohol (Vadlamudi et al., 2004). Withdrawal symptoms pre- more, OT has a calming effect by increasing parasympa- cipitated by picrotoxin treatment of alcohol-dependent mice thetic activity (Michelini, 2007; Higa-Taniguchi et al., 2009). or naloxone treatment of morphine-dependent mice were In human studies, acute intranasal OT administration markedly reduced by OT treatment (Szabó et al., 1987). decreased anticipatory stress-induced anxiety (de Oliveira et al., 2012) and had synergistic effects with social support 4.1. Oxytocin blockade of withdrawal in alcohol dependent in reducing stress-induced anxiety (Heinrichs et al., 2003). In patients fMRI studies, intranasal OT reduced amygdala activation by various stressful stimuli in men but had contrasting effects To test whether OT inhibits alcohol withdrawal in humans, on amygdala activation by different stressful stimuli in Pedersen et al. (2013) conducted a pilot randomized, double- women (Zink and Meyer-Lindenberg, 2012). Intranasal OT blind 3-day trial of BID intranasal OT vs. placebo in 11 treatment also decreases cortisol release during and follow- alcohol-dependent inpatient subjects undergoing detoxifica- ing psychosocial stressors (Heinrichs et al., 2003) and tion with symptom-driven PRN (as needed) lorazepam. Indi- increases heart rate variability, an index of parasympathetic viduals were recruited by advertising and evaluated at an activity (Norman et al., 2011; Kemp et al., 2012). OT receptors outpatient visit where informed consent, medical, psychiatric are expressed in cultured human monocytes and macro- and drinking history were obtained and blood and urine phages and OT suppresses release of proinflammatory collected for laboratory tests. Subjects were recruited who 114 brain research 1580 (2014) 102–123

had been drinking heavily (8–30 standard drinks/day) for at 21 least 2 weeks prior to recruitment, had previously experi- enced 2 or more days of debilitating withdrawal symptoms, 18 but not delirium tremens or seizures, after a heavy drinking period, and had no unstable medical or psychiatric disorders 16 based on interview and laboratory tests. They were admitted to an inpatient research unit. After obtaining baseline alcohol withdrawal symptom ratings using a modified (vital signs 13 were included in the score) Clinical Institute Withdrawal Assessment of Alcohol Scale (CIWA, Sullivan et al., 1989), 11 subjects received their first intranasal dose of the substance a to which they had been randomized; OT (24 IU/dose) or 8 placebo. They received intranasal test doses BID for 3 days. CIWA scores were obtained every 4 h throughout the trial (mg) Lorazepam Total 5 except when subjects slept at night or whenever subjects reported or research nurses observed an increase in with- 3 drawal symptoms. Lorazepam (2 mg) was given orally when- ever CIWA scores were Z12. Another CIWA score was obtained 1 h after each lorazepam dose. Lorazepam doses 0 Oxytocin Placebo were repeated if CIWA scores remained Z10. Subjects rated themselves on the Penn Alcohol Craving Scale (PACS, Fig. 7 – Mean ( SD) total lorazepam (mg) required to þ Flannery et al., 1999) and the Profile of Mood States (POMS, complete medical detoxification from alcohol during 3 days McNair et al., 1971) midday on admission days 2 and 3. of intranasal oxytocin or placebo treatment. a Denotes The mean number of standard drinks per day prior to p .0015 compared to placebo. ¼ enrollment did not differ between the OT (N 7) and placebo ¼ (N 4) groups (17.776.0 [SD] vs.14.571.7). The placebo group 16 ¼ Oxytocin compared to the OT group required 5 times more total 14 lorazepam to complete detoxification (Fig. 7). Mean CIWA Placebo 12 scores were almost 3 times higher in the placebo compared to 10 the OT group on admission days 1 and 2 (Fig. 8) but not on admission day 3. POMS Tension/Anxiety subscores were 8 ba significantly lower on admission day 2 in the OT compared 6 to the placebo group (10.376.0 vs. 23.876.7, p .007) but did ¼ 4 not differ significantly on admission day 3. PACS scores did not differ between treatment groups on admission day 2 or 3. Score CIWA Daily Mean 2 Placebo recipients, all of whom had been drinking r16 mean 0 Day 1 Day 2 Day 3 drinks/day (N 4), required an average total of 16.574.4 mg of ¼ lorazepam to complete detoxification. Oxytocin recipients Fig. 8 – Mean ( SD) daily CIWA scores on inpatient þ who consumed r16 mean drinks/day (N 4), required no admission days 1–3 in subjects receiving intranasal oxytocin ¼ lorazepam (p .005). CIWA scores were significantly lower on or placebo. a Denotes p .0015 and b denotes po.0001 ¼ ¼ all 3 admission days in this OT subgroup. Oxytocin treatment compared to placebo. was not associated with side effects or clinically significant changes in laboratory values. This is the first evidence that OT may potently block alcohol withdrawal in humans. consists of psychosocial interventions including 12-step pro- grams and motivational interviewing though perhaps less 4.2. Questions to guide future research on oxytocin than 25% of patients with alcohol dependence receive treat- mechanisms in alcohol and other dependence disorders ment (Hasin and Grant, 2004). Relatively recently, the FDA approved two drugs that have exhibited some efficacy in 4.2.1. Will oxytocin treatment decrease drinking and craving increasing partial or complete abstinence (Kranzler and Van in alcohol-dependent subjects? Kirk, 2001): naltrexone, a μ-opioid receptor antagonist, and The Center for Disease Control places alcohol as the 3rd acamprosate, thought to modulate N-methyl-D-aspartate recep- cause of preventable deaths following nicotine use and over- tors. However, the use of these medications for alcohol depen- weight (Centers for Disease Control and Prevention, 2009). dence has been limited (Mark et al., 2009)becausebothhave The economic costs of alcohol misuse for 2006 are estimated small to medium effect sizes (Kranzler and Van Kirk, 2001). at $223 billion/year (Bouchery et al., 2011). Epidemiological Furthermore, it is suggested (FDA label) that patients be absti- studies show that the 12-month prevalence rate for DSM-IV nent from drinking prior to starting naltrexone or acamprosate alcohol dependence in USA is 5–6% for men and 2–3% for which adds another potential barrier in the clinical setting. women (Grant et al., 2004). Despite these grim statistics, For several reasons, OT has a potentially attractive profile treatment options are limited. Currently, treatment to for the treatment of alcohol-dependent individuals. First, decrease drinking in alcohol dependent individuals primarily numerous animal and human studies consistently show that brain research 1580 (2014) 102–123 115

OT is anxiolytic and reduces fear (McCarthy et al., 1996; Windle relapse: persistent craving for alcohol; high anxiety; decreased et al., 1997; Bale et al., 2001; Mantella et al., 2003; Heinrichs ability to cope with stress; and consumption of large quantities et al., 2003; Amico et al., 2004; Ring et al., 2006; Labuschagne of alcohol upon relapse. If OT treatment diminishes - et al., 2010; de Oliveira et al., 2012) including in the context of hypnotic tolerance, patients may be less likely to relapse after alcohol detoxification (see above) and in an animal model of medical detoxification. This prospect is supported by the relapse (see below). Anxiety is hypothesized to impede recov- findings of Myrick et al. (2009) that treatment of alcohol ery and contribute to relapse (Kushner et al., 2000). In clinical withdrawal with gabapentin, a non-sedative-hypnotic drug, and laboratory studies, alcohol has been shown to have both was as effective as lorazepam but resulted in less alcohol an acute anxiolytic effect and a delayed anxiogenic effect. This consumption after completion of detoxification. There are led Kushner et al. (2000) to propose a feed-forward cycle where potential drawbacks to OT treatment of alcohol dependence. the initial anxiolytic effect of alcohol is followed by an Rapid reduction of tolerance by OT could increase the vulner- anxiogenic effect which then promotes the use of alcohol ability of heavy drinkers to toxic effects if they relapse and because of its anxiolytic effect. As shown in the clinical trial consume large amounts of alcohol. Also, it is unknown described above (Pedersen et al., 2013), OT potently reduces whether OT decreases risk of developing delirium tremens other withdrawal symptoms as well. Dependent individuals and seizures and should not be substituted for benzodiaze- are often unable to refrain from drinking because withdrawal pines, which are known to reduce these risks, in the treatment symptoms become unbearable. In the withdrawal treatment of patients with history of dangerous withdrawal. trial, intranasal OT was administered safely to subjects who had been drinking heavily until admission to the research 4.2.3. Does OT block withdrawal and decrease consumption unit. This indicates that a period of abstinence from alcohol of other addictive substances? may not be necessary prior to initiating OT treatment. Craving Reports that OT inhibits tolerance formation to opioids and for alcohol, like withdrawal, may be related to tolerance cocaine, reduces opioid withdrawal and exerts other anti- formation. If OT blocks withdrawal by reversing neuroadapta- addictive effects in animals (Sarnyai and Kovács, 1994; tion underlying tolerance, it may decrease craving by a similar Kovács et al., 1998; Yang et al., 2010; McGregor and Bowen, mechanism. OT treatment also blocked increased alcohol 2012) suggest that OT may be effective in treating a wide intake in P (alcohol-preferring) rats subjected to repeated range of substance dependence disorders (McGregor and alcohol deprivation combined with restraint stress, an animal Bowen, 2012). Clinical trials testing OT on withdrawal from model of craving and relapse (Breese et al., 2004; Overstreet and consumption of opioids and cocaine as well as other et al., 2007; Knapp et al., 2010). These considerations support addictive substances are of great importance. Demonstrating the hypothesis that intranasal OT may decrease drinking and that OT is an effective treatment for multiple substance craving in alcohol-dependent individuals. dependence disorders would be a major advance in addiction medicine and would inspire new research to determine if 4.2.2. Does oxytocin inhibit tolerance formation and reverse CNS OT is involved in pathophysiological mechanisms com- established tolerance in human subjects? mon to many if not all addictions. The repeated demonstration in animals that OT inhibits formation of tolerance to alcohol suggests that OT may block alcohol withdrawal by reversing tolerance. This hypothesis 5. Speculation on the evolutionary origins should be tested by comparing the effects of intranasal OT of oxytocin efficacy in schizophrenia and alcohol administration on responses to repeated ingestion of a withdrawal standard dose of alcohol in normal volunteers and alcohol- dependent subjects. The hypothesis would be supported if OT In Article 2 (in this volume), Pedersen et al. argue that the compared to placebo inhibits decline in the magnitude of selection of OT to activate avid and sustained maternal ethanol effects in normals and increases sensitivity to etha- behavior was essential for the evolution of placental mam- nol effects in dependent individuals. Demonstrating that OT mals. Research in rodents and sheep indicate that in the sex blocks tolerance formation in humans would reinvigorate hormone milieu of late pregnancy and parturition OT facil- investigation of the underlying mechanisms as well as the itates the onset of maternal behavior by enabling activation relationship of CNS OT to other neurochemical systems that of reward pathways by stimuli from newborns and by have been implicated in alcohol dependence (Wee and Koob, expression of hard-wired, species-specific nurturing beha- 2010; Pava and Woodward, 2012; Roberto et al., 2012; Lovinger viors (Numan and Insel, 2003; Pedersen, 2013). However, OT and Roberto, 2013). may have been selected to play other roles that were This line of research may also have important clinical important for the considerably higher quality of maternal implications. OT could have advantages over benzodiaze- behavior that was necessary for the evolution of placental pines, the currently recommended treatment for alcohol mammals. Rapid and accurate perception of sensory cues withdrawal symptoms. While benzodiazepines are very from newborns and timely expression of appropriate care- effective and relatively safe (Mayo-Smith, 1997; Daeppen taking behaviors in response to those cues may have required et al., 2002), they are, like alcohol, sedative- and are an up-grade of sensory-motor gating systems that enabled known to have addictive properties and to enhance the mothers to effectively nurture their offspring but also depressant effects of alcohol. Furthermore, benzodiazepine endowed placental mammalian brains in general with treatment may maintain elevated sedative-hypnotic tolerance greater ability in this area. If neural systems were selected which could contribute to factors that increase vulnerability to through which OT enhances sensory-motor gating in 116 brain research 1580 (2014) 102–123

mammals, especially mothers, perhaps it is through those Arnsten, A.F., 2009. Stress signaling pathways that impair neural systems that OT reverses deficits in PPI in animal prefrontal cortex structure and function. Nat. Rev. Neurosci. models of psychosis and contributes to OT reduction of 10, 410–422. Averbeck, B.B., Bobin, T., Evans, S., Shergill, S.S., 2012. Emotion psychotic symptoms in schizophrenia. A potential test of this recognition and oxytocin in patients with schizophrenia. hypothesis in rodents would be to compare PPI in new Psychol. Med., 1–8. mothers and nulliparous females and examine the role of Bakharev, V.D., Tikhomirov, S.M., Lozhkina, T.K., 1986. endogenous OT in differences that may be discovered. Psychotropic properties of oxytocin. Neurosci. Behav. Physiol. The success of mammalian mothering behavior requires 16, 160–164. that maternal care be sustained over extended periods until Bale, T.L., Davis, A.M., Auger, A.P., Dorsa, D.M., McCarthy, M.M., offspring can survive independently. Similarly, monogamous 2001. CNS region-specific oxytocin receptor expression: relationships in species such as the prairie vole require that importance in regulation of anxiety and sex behavior. J. Neurosci. 21, 2546–2552. selective attachments are maintained over periods as long as Bale, T.L., Pedersen, C.A., Dorsa, D.M., 1995. CNS oxytocin the lifetimes of the pair-bonded couples. Perpetuation of receptor mRNA expression and regulation by gonadal steroids. these reproductively critical attachments may have required Adv. Exp. Med. Biol. 395, 269–280. the evolution of neural mechanisms by which OT prevents Bales, K.L., Abdelnabi, M., Cushing, B.S., Ottinger, M.A., Carter, C.S., habituation to the rewarding stimuli from offspring or pair- 2004. Effects of neonatal oxytocin manipulations on male bonded mates. These hypothesized mechanisms may under- reproductive potential in prairie voles. Physiol. Behav. 81, lie the efficacy of OT in reducing tolerance formation to 519–526. Bales, K.L., Carter, C.S., 2003a. Developmental exposure to addictive substances, all of which activate reward pathways. oxytocin facilitates partner preferences in male prairie voles However, while central administration of OT antagonists (Microtus ochrogaster). Behav. Neurosci. 117, 854–859. potently disrupt the onset of maternal behavior acute ICV Bales, K.L., Carter, C.S., 2003b. Sex differences and developmental injection of OT antagonists only weakly inhibits some com- effects of oxytocin on aggression and social behavior in prairie ponents of established maternal behavior in lactating rats voles (Microtus ochrogaster). Horm. Behav. 44, 178–184. (Numan and Insel, 2003; Pedersen, 2013), evidence that does Bales, K.L., van Westerhuyzen, J.A., Lewis-Reese, A.D., Grotte, N.D., not support a role for central OT in preventing decline in Lanter, J.A., Carter, C.S., 2007. Oxytocin has dose-dependent maternal responsiveness to offspring. The effects of central developmental effects on pair-bonding and alloparental care in female prairie voles. Horm. Behav. 52, 274–279. OT antagonist administration on established pair bonds have Baskerville, T.A., Douglas, A.J., 2010. Dopamine and oxytocin yet to be tested. It may require OT antagonist treatment for interactions underlying behaviors: potential contributions to longer periods to adequately test whether central OT plays a behavioral disorders. CNS Neurosci. Ther. 16, e92–e123. significant role in maintaining established maternal behavior Baumgartner, T., Heinrichs, M., Vonlanthen, A., Fischbacher, U., or monogamous pair bonds. Alternatively, the hypothesized Fehr, E., 2008. Oxytocin shapes the neural circuitry of trust role of OT in blocking habituation to stimuli from significant and trust adaptation in humans. Neuron 58, 639–650. others may have evolved in mammalian species other than Beckmann, H., Lang, R.E., Gattaz, W.F., 1985. Vasopressin– oxytocin in cerebrospinal fluid of schizophrenic patients and rodents in which long-term maintenance of social bonds normal controls. Psychoneuroendocrinology 10, 187–191. fi increased reproductive tness. Beitchman, J.H., Zai, C.C., Muir, K., Berall, L., Nowrouzi, B., Choi, In the ways discussed above, recent discoveries of clinical E., Kennedy, J.L., 2012. Childhood aggression, callous- efficacy of OT in schizophrenia and alcohol withdrawal may unemotional traits and oxytocin genes. Eur. Child Adolesc. provide new insights into OT mechanisms which may have Psychiatry 21, 125–132. been selected for during the evolution of placental mammals Bellack, A.S., Schooler, N.R., Marder, S.R., Kane, J.M., Brown, C.H., to facilitate maternal-infant and other social attachments. Yang, Y., 2004. Do clozapine and risperidone affect social competence and problem solving?. Am. J. Psychiatry 161, 364–367. Bernstein, H.G., Stanarius, A., Baumann, B., Henning, H., Krell, D., Acknowledgments Danos, P., Falkai, P., Bogerts, B., 1998. Nitric oxide synthase- containing neurons in the human hypothalamus: reduced Many thanks to David Penn and James C. Garbutt for their number of immunoreactive cells in the paraventricular invaluable contributions to the execution of our OT clinical nucleus of depressive patients and schizophrenics. trials summarized in this article and to their helpful sugges- Neuroscience 83, 867–875. tions about earlier versions of this manuscript. Bethlehem, R.A.I., van Honk, J., Auyeung, B., Baron-Cohen, S., 2013. Oxytocin, brain physiology, and functional connectivity: a review of intranasal oxytocin fMRI studies. references Psychoneuroendocrinology 38, 962–974. Boccia, M.L., Petrusz, P., Suzuki, K., Marson, L., Pedersen, C.A., 2013. Immunohistochemical localization of oxytocin receptors Adolphs, R., Tranel, D., Damasio, A.R., 1998. The human amygdala in human brain. Neuroscience 253, 155–164. in social judgment. Nature 393, 470–474. Bora, E., Yucel, M., Pantelis, C., 2009. Theory of mind impairment Amico, J.A., Mantella, R.C., Vollmer, R.R., Li, X., 2004. Anxiety and in schizophrenia: meta-analysis. Schizophr. Res. 109, 1–9. stress responses in female oxytocin deficient mice. J. Born, J., Lange, T., Kern, W., McGregor, G.P., Bickel, U., Fehm, H.L., Neuroendocrinol. 16, 319–324. 2002. Sniffing neuropeptides: a transnasal approach to the Andari, E., Duhamel, J.R., Zalla, T., Herbrecht, E., Leboyer, M., human brain. Nat. Neurosci. 5, 514–516. Sirigu, A., 2010. Promoting social behavior with oxytocin in Bouchery, E.E., Harwood, H.J., Sacks, J.J., Simon, C.J., Brewer, R.D., high-functioning autism spectrum disorders. Proc. Natl. Acad. 2011. Economic costs of excessive alcohol consumption in the Sci. USA 107, 4389–4394. U.S., 2006. Am. J. Prev. Med. 41, 516–524. brain research 1580 (2014) 102–123 117

Braff, D., Geyer, M., Swerdlow, N., 2001. Human studies of Choleris, E., Devidze, N., Kavaliers, M., Pfaff, D.W., 2008. Steroidal/ prepulse inhibition of startle: normal subjects, patient groups, neuropeptide interactions in hypothalamus and amygdala and pharmacological studies. Psychopharmacology 156, related to social anxiety. Prog. Brain Res. 170, 291–303. 234–258. Clodi, M., Vila, G., Geyeregger, R., Riedl, M., Stulnig, T.M., Struck, J., Breese, G.R., Knapp, D.J., Overstreet, D.H., 2004. Stress Luger, T.A., Luger, A., 2008. Oxytocin alleviates the sensitization of ethanol withdrawal-induced reduction in neuroendocrine and cytokine response to bacterial endotoxin social interaction: inhibition by CRF-1 and benzodiazepine in healthy men. Am. J. Physiol. Endocr. Metab. 295, E686–E691. receptor antagonists and a 5-HT1A-receptor agonist. Cohen, H., Kaplan, Z., Kozlovsky, N., Gidron, Y., Matar, M.A., Neuropsychopharmacology 29, 470–482. Zohar, J., 2010. Hippocampal microinfusion of oxytocin Broome, M.R., Woolley, J.B., Tabraham, P., Johns, L.C., Bramon, E., attenuates the behavioural response to stress by means of Murray, G.K., Pariante, C., McGuire, P.K., Murray, R.M., 2005. dynamic interplay with the glucocorticoid–catecholamine What causes the onset of psychosis?. Schizophr. Res. 79, responses. J. Neuroendocrinol. 22, 889–904. 23–34. Corcoran, C., Whitaker, A., Coleman, E., Fried, J., Feldman, J., Brown, A.S., 2011. The environment and susceptibility to Goudsmit, N., Malaspina, D., 2005. Olfactory deficits, cognition schizophrenia. Prog. Neurobiol. 93, 23–58. and negative symptoms in early onset psychosis. Schizophr. Bruins, J., Hijman, R., Van Ree, J.M., 1992. Effect of a single dose of Res. 80, 283–293. Couture, S.M., Penn, D.L., Roberts, D.L., 2006. The functional des-glycinamide-[Arg8]vasopressin or oxytocin on cognitive significance of social cognition in schizophrenia: a review. processes in young healthy subjects. Peptides 13, 461–468. Schizophr. Bull. 32 (Suppl. 1), S44–S63. Bru¨ ne, M., 2003. Social cognition and behaviour in schizophrenia. Czyrak, A., Mackowiak, M., Chocyk, A., Fijal, K., Wedzony, K., In: Bru¨ ne, M., Ribbert, H., Schiefenho¨vel, W. (Eds.), The Social 2003. Role of glucocorticoids in the regulation of dopaminergic Brain. John Wiley & Sons, Ltd., Chichester, pp. 277–313. neurotransmission. Pol. J. Pharmacol. 55, 667–674. Buchanan, R.W., Breier, A., Kirkpatrick, B., Ball, P., Carpenter Jr., W.T., Daeppen, J.B., Gache, P., Landry, U., Sekera, E., Schweizer, V., 1998. Positive and negative symptom response to clozapine in Gloor, S., Yersin, B., 2002. Symptom-triggered vs. fixed- schizophrenic patients with and without the deficit syndrome. schedule doses of benzodiazepine for alcohol withdrawal: a Am. J. Psychiatry 155, 751–760. randomized treatment trial. Arch. Intern. Med. 162, 1117–1121. Buchanan, R.W., Javitt, D.C., Marder, S.R., Schooler, N.R., Gold, J.M., Dallman, M.F., Akana, S.F., Strack, A.M., Scribner, K.S., Pecoraro, McMahon, R.P., Heresco-Levy, U., Carpenter, W.T., 2007. The N., La Fleur, S.E., Houshyar, H., Gomez, F., 2004. Chronic stress- Cognitive and Negative Symptoms in Schizophrenia Trial induced effects of corticosterone on brain: direct and indirect. (CONSIST): the efficacy of glutamatergic agents for negative Ann. N. Y. Acad. Sci. 1018, 141–150. symptoms and cognitive impairments. Am. J. Psychiatry 164, Daskalakis, Z.J., Christensen, B.K., Fitzgerald, P.B., Chen, R., 2008. 1593–1602. Dysfunctional neural plasticity in patients with Bujanow, W., 1972. Hormones in the treatment of psychoses. Br. schizophrenia. Arch. Gen. Psychiatry 65, 378–385. Med. J. 4, 298. Davis, M.H., 1983. Measuring individual differences in empathy: Buske-Kirschbaum, A., Jobst, S., Wustmans, A., Kirschbaum, C., evidence for a multidimensional approach. J. Personal. Soc. Rauh, W., Hellhammer, D., 1997. Attenuated free cortisol Psychol. 44, 113–126. response to psychosocial stress in children with atopic de Oliveira, D.C., Zuardi, A.W., Graeff, F.G., Queiroz, R.H., Crippa, J. dermatitis. Psychosom. Med. 59, 419–426. A., 2012. Anxiolytic-like effect of oxytocin in the simulated Cahn, W., van Haren, N.E., Hulshoff Pol, H.E., Schnack, H.G., public speaking test. J. Psychopharmacol. 26, 497–504. Caspers, E., Laponder, D.A., Kahn, R.S., 2006. Brain volume Delis, D., Kramer, J., Kaplan, E., Ober, B., 2000. In: California Verbal changes in the first year of illness and 5-year outcome of Learning Test. The Psychological Corporation, Harcourt schizophrenia. Br. J. Psychiatry 189, 381–382. Assessment Company, San Antonio, TX. Caldwell, H.K., Stephens, S.L., Young, W.S., 2009. Oxytocin as a den Boer, J.A., Westenberg, H.G., 1992. Oxytocin in obsessive natural antipsychotic: a study using oxytocin knockout mice. compulsive disorder. Peptides 13, 1083–1085. Mol. Psychiatry 14 (2), 190–196. Dhakar, M.B., Rich, M.E., Reno, E.L., Lee, H.J., Caldwell, H.K., 2012. Carlsson, A., 1977. Does dopamine play a role in schizophrenia?. Heightened aggressive behavior in mice with lifelong versus Psychol. Med. 7, 583–597. postweaning knockout of the oxytocin receptor. Horm. Behav. Carlsson, A., Carlsson, M.L., 2006. A dopaminergic deficit 62, 86–92. hypothesis of schizophrenia: the path to discovery. Dialogues Domes, G., Heinrichs, M., Michel, A., Berger, C., Herpertz, S.C., Clin. Neurosci. 8, 137–142. 2007a. Oxytocin improves “mind-reading” in humans. Biol. Carter, C.S., Lederhendler, I.I., Kirkpatrick, B., 1999. In: The Psychiatry. 61, 731–733. Integrative Neurobiology of Affiliation. The MIT Press, Domes, G., Heinrichs, M., Glascher, J., Buchel, C., Braus, D.F., Cambridge, MA. Herpertz, S.C., 2007b. Oxytocin attenuates amygdala Casey, B.J., Jones, R.M., Levita, L., Libby, V., Pattwell, S.S., Ruberry, responses to emotional faces regardless of valence. Biol. E.J., Soliman, F., Somerville, L.H., 2010. The storm and stress of Psychiatry. 62, 1187–1190. adolescence: insights from human imaging and mouse Domes, G., Lischke, A., Berger, C., Grossmann, A., Hauenstein, K., genetics. Dev. Psychobiol. 52, 225–235. Heinrichs, M., Herpertz, S.C., 2010. Effects of intranasal Cavus, I., Reinhart, R.M., Roach, B.J., Gueorguieva, R., Teyler, T.J., oxytocin on emotional face processing in women. Clapp, W.C., Ford, J.M., Krystal, J.H., Mathalon, D.H., 2012. Psychoneuroendocrinology 35, 83–93. Impaired visual cortical plasticity in schizophrenia. Biol. Doty, R.L., Shaman, P., Kimmelman, C.P., Dann, M.S., 1984. University of Pennsylvania Smell Identification Test: a rapid Psychiatry 71, 512–520. Centers for Disease Control and Prevention, 2009. FastStats, quantitative olfactory function test for the clinic. Laryngoscope 94, 176–178. Alcohol Use. Available at: 〈http://www.cdc.gov/nchs/fastats/ Ebstein, R.P., Knafo, A., Mankuta, D., Chew, S.H., Lai, P.S., 2012. alcohol.htm〉 (cf. updated 22.10.09). The contributions of oxytocin and vasopressin pathway genes Chen, Q., Schreiber, S.S., Brinton, R.D., 2000. Vasopressin and to human behavior. Horm. Behav. 61, 359–379. oxytocin receptor mRNA expression during rat telencephalon Emiliano, A.B., Cruz, T., Pannoni, V., Fudge, J.L., 2007. The development. Neuropeptides 34, 173–180. interface of oxytocin-labeled cells and serotonin transporter- 118 brain research 1580 (2014) 102–123

containing fibers in the primate hypothalamus: a substrate for schizophrenic patients does not differ from control subjects SSRIs therapeutic effects?. Neuropsychopharmacology 32, and is not changed by neuroleptic medication. Schizophr. Res. 977–988. 11, 273–276. Epperson, C.N., McDougle, C.J., Price, L.H., 1996. Intranasal Goldman, L., Winget, C., Hollingshead, G., Levine, S., 1973. oxytocin in obsessive-compulsive disorder. Biol. Psychiatry 40, Postweaning development of negative feedback in the 547–549. pituitary-adrenal system of the rat. Neuroendocrinology 12, Fair, D.A., Dosenbach, N.U., Church, J.A., Cohen, A.L., Brahmbhatt, 199–211. S., Miezin, F.M., Barch, D.M., Raichle, M.E., Petersen, S.E., Goldman, M., Marlow-O’Connor, M., Torres, I., Carter, C., 2008. Schlaggar, B.L., 2007. Development of distinct control Diminished plasma oxytocin in schizophrenic patients with networks through segregation and integration. Proc. Natl. neuroendocrine dysfunction and emotional deficits. Acad. Sci. USA 104, 13507–13512. Schizophr. Res. 98, 247–255. Fehm-Wolfsdorf, G., Born, J., Voigt, K.H., Fehm, H.L., 1984. Human Goldman, M.B., Gomes, A.M., Carter, C.S., Lee, R., 2011. Divergent memory and neurohypophyseal hormones: opposite effects of effects of two different doses of intranasal oxytocin on facial vasopressin and oxytocin. Psychoneuroendocrinology 9, affect discrimination in schizophrenic patients with and 285–292. without polydipsia. Psychopharmacology (Berl.) 216, 101–110. Fehm-Wolfsdorf, G., Born, J., 1991. Behavioral effects of Goto, Y., Yang, C.R., Otani, S., 2010. Functional and dysfunctional neurohypophyseal peptides in healthy volunteers: 10 years of synaptic plasticity in prefrontal cortex: roles in psychiatric research. Peptides 12, 1399–1406. disorders. Biol. Psychiatry 67, 199–207. Feifel, D., Reza, T., 1999. Oxytocin modulates psychotomimetic- Grace, A.A., 2004. Developmental dysregulation of the dopamine induced deficits in sensorimotor gating. Psychopharmacology system and the pathophysiology of schizophrenia. In: (Berl.) 141, 93–98. Keshavan, M.S., Kennedy, J.L., Murray, R. (Eds.), Feifel, D., MacDonald, K., Nguyen, A., Cobb, P., Warlan, H., Neurodevelopment and Schizophrenia. Cambridge University Galangue, B., Minassian, A., Becker, O., Cooper, J., Perry, W., Press, Cambridge, pp. 273–294. Lefebvre, M., Gonzales, J., Hadley, A., 2010. Adjunctive Grant, B.F., Stinson, F.S., Dawson, D.A., Chou, S.P., Dufour, M.C., intranasal oxytocin reduces symptoms in schizophrenia Compton, W., Pickering, R.P., Kaplan, K., 2004. Prevalence and patients. Biol. Psychiatry 68, 678–680. co-occurrence of substance use disorders and independent Feifel, D., MacDonald, K., Cobb, P., Minassian, A., 2012. Adjunctive mood and anxiety disorders: results from the National intranasal oxytocin improves verbal memory in people with Epidemiologic Survey on Alcohol and Related Conditions. schizophrenia. Schizophr. Res. 139, 207–210. Arch. Gen. Psychiatry 61, 807–816. Fenigstein, A., Vanable, P.A., 1992. Paranoia and self- Green, M.F., Olivier, B., Crawley, J.N., Penn, D.L., Silverstein, S., consciousness. J. Personal. Soc. Psychol. 62, 129–138. 2005. Social cognition in schizophrenia: recommendations Ferrier, B.M., Kennett, D.J., Devlin, M.C., 1980. Influence of from the measurement and treatment research to improve oxytocin on human memory processes. Life Sci. 27, 2311–2317. cognition in schizophrenia new approaches conference. Fett, A.K, Gracia-Dominguez, M., Viechtbauer, W., Penn, D.L., van Schizophr. Bull. 31, 882–887. Os, J., Krabbendam, L., 2011. A meta-analytic study on Gregory, S., Connelly, J., Towers, A., Johnson, J., Biscocho, D., cognition, social cognition, and functional outcome in non- Markunas, C., Lintas, C., Abramson, R., Wright, H., Ellis, P., affective psychosis. Neurosci. Biobehav. Rev. 35, 573–588. 2009. Genomic and epigenetic evidence for oxytocin receptor Finlay, J.M., Zigmond, M.J., 1997. The effects of stress on central deficiency in autism. BMC Med. 7, 62. dopaminergic neurons: possible clinical implications. Grippo, A.J., Trahanas, D.M., Zimmerman II, R.R., Porges, S.W., Neurochem. Res. 22, 1387–1394. Carter, C.S., 2009. Oxytocin protects against negative Fischer-Shofty, M., Shamay-Tsoory, S.G., Harari, H., Levkovitz, Y., behavioral and autonomic consequences of long-term social 2010. The effect of intranasal administration of oxytocin on fear recognition. Neuropsychologia 48, 179–184. isolation. Psychoneuroendocrinology 34, 1542–1553. Fitzsimmons, J., Kubicki, M., Shenton, M.E., 2013. Review of Groppe, S.E., Gossen, A., Rademacher, L., Hahn, A., Westphal, L., functional and anatomical brain connectivity findings in Gru¨ nder, G., Spreckelmeyer, K.N., 2013. Oxytocin influences schizophrenia. Curr. Opin. Psychiatry 26, 172–187. processing of socially relevant cues in the ventral tegmental Flannery, B.A., Volpicelli, J.R., Pettinati, H.M., 1999. Psychometric area of the human brain. Biol. Psychiatry 74 (3), 172–179. properties of the Penn Alcohol Craving Scale. Alcohol. Clin. Guastella, A.J., MacLeod, C., 2012. A critical review of the Exp. Res. 23, 1289–1295. influence of oxytocin nasal spray on social cognition in Galvan, A., Hare, T.A., Parra, C.E., Penn, J., Voss, H., Glover, G., humans: evidence and future directions. Horm. Behav. 61 (3), Casey, B.J., 2006. Earlier development of the accumbens 410–418. relative to orbitofrontal cortex might underlie risk-taking Guastella, A.J., Mitchell, P.B., Dadds, M.R., 2008. Oxytocin behavior in adolescents. J. Neurosci. 26, 6885–6892. increases gaze to the eye region of human faces. Biol. Ganzel, B.L., Eckenrode, J.J., Kim, P., Wethington, E., Horowitz, E., Psychiatry 63, 3–5. Temple, E., 2007. Salivary cortisol levels and mood vary by Guastella, A.J., Einfeld, S.L., Gray, K.M., Rinehart, N.J., Tonge, B.J., lifetime trauma exposure in a sample of healthy women. J. Lambert, T.J., Hickie, I.B., 2010. Intranasal oxytocin improves Trauma. Stress 20, 689–699. emotion recognition for youth with autism spectrum Geyer, M.A., Krebs-Thomson, K., Braff, D.L., Swerdlow, N.R., 2001. disorders. Biol. Psychiatry 67, 692–694. Pharmacological studies of prepulse inhibition models of Guy, W., Bonato, R., 1970. In: Manual for the ECDEU Assessment sensorimotor gating deficits in schizophrenia: a decade in Battery 2nd revised ed. National Institute of Mental Health, review. Psychopharmacology (Berl.) 156, 117–154. Chevy Chase, MD. Geyer, M.A., Ellenbroek, B., 2003. Animal behavior models of the Hammock, E.A., Young, L.J., 2006. Oxytocin, vasopressin and pair mechanisms underlying antipsychotic atypicality. Prog. bonding: implications for autism. Philos. Trans. R. Soc. Lond. Neuropsychopharmacol. Biol. Psychiatry 27, 1071–1079. B: Biol. Sci. 361, 2187–2198. Gimpl, G., Fahrenholz, F., 2001. The oxytocin receptor system: Harvey, P.D., Patterson, T.L., Potter, L.S., Zhong, K., Brecher, M., structure, function, and regulation. Physiol. Rev. 81, 629–683. 2006. Improvement in social competence with short-term Glovinsky, D., Kalogeras, K.T., Kirch, D.G., Suddath, R., Wyatt, R.J., atypical antipsychotic treatment: a randomized, double-blind 1994. Cerebrospinal fluid oxytocin concentration in comparison of quetiapine versus risperidone for social brain research 1580 (2014) 102–123 119

competence, social cognition, and neuropsychological Kelly, A., Uddin, L.Q., Biswal, B.B., Castellanos, F.X., Milham, M.P., functioning. Am. J. Psychiatry 163, 1918–1925. 2008. Competition between functional brain networks Hasan, A., Nitsche, M.A., Rein, B., Schneider-Axmann, T., Guse, B., mediates behavioral variability. Neuroimage 39, 527–537. Gruber, O., Falkai, P., Wobrock, T., 2011. Dysfunctional long- Kemp, A.H., Quintana, D.S., Kuhnert, R.-L., Griffiths, K., Hickie, I. term potentiation-like plasticity in schizophrenia revealed by B., Guastella, A.J., 2012. Oxytocin increases heart rate transcranial direct current stimulation. Behav. Brain Res. 224, variability in humans at rest: implications for social approach- 15–22. related motivation and capacity for social engagement. PloS Hasin, D.S., Grant, B.F., 2004. The co-occurrence of DSM-IV alcohol One 7, e44014. abuse in DSM-IV alcohol dependence: results of the National Kenny, F.M., Preeyasombat, C., Migeon, C.J., 1966. Cortisol Epidemiologic Survey on Alcohol and Related Conditions on production rate. II. Normal infants, children, and adults. heterogeneity that differ by population subgroup. Arch. Gen. Pediatrics 37, 34–42. Psychiatry 61, 891–896. Keri, S., Kiss, I., Kelemen, O., 2009. Sharing secrets: oxytocin and Heinrichs, M., Baumgartner, T., Kirschbaum, C., Ehlert, U., 2003. trust in schizophrenia. Soc. Neurosci. 4, 287–293. Social support and oxytocin interact to suppress cortisol and Kirkpatrick, B., Kopelowicz, A., Buchanan, R.W., Carpenter Jr., W. subjective responses to psychosocial stress. Biol. Psychiatry T., 2000. Assessing the efficacy of treatments for the deficit 54, 1389–1398. syndrome of schizophrenia. Neuropsychopharmacology 22, Heinrichs, M., Meinlschmidt, G., Wippich, W., Ehlert, U., 303–310. Hellhammer, D.H., 2004. Selective amnesic effects of oxytocin Kirsch, P., Esslinger, C., Chen, Q., Mier, D., Lis, S., Siddhanti, S., on human memory. Physiol. Behav. 83, 31–38. Gruppe, H., Mattay, V.S., Gallhofer, B., Meyer-Lindenberg, A., Higa-Taniguchi, K.T., Felix, J.V., Michelini, L.C., 2009. Brainstem 2005. Oxytocin modulates neural circuitry for social cognition oxytocinergic modulation of heart rate control in rats: effects and fear in humans. J. Neurosci. 25, 11489–11493. of hypertension and exercise training. Exp. Physiol. 94, Knapp, D.J., Overstreet, D.H., Hubka, J.C., Breese, G.R., Pedersen, C.A., 1103–1113. 2010. Inhibitory effects of oxytocin on anxiety-like behavior and Hoschl, C., Hajek, T., 2001. Hippocampal damage mediated by alcohol consumption in P rats subjected to repeated cycles of – a neuropsychiatric research challenge. Eur. voluntary ethanol consumption and stress. In: Proceedings of the 33rd Annual Research Society on Alcoholism Scientific Arch. Psychiatry Clin. Neurosci. 251 (Suppl. 2), S1181–S1188. Hrabovszky, E., Liposits, Z., 2008. Novel aspects of glutamatergic Meeting, vol. 145A. San Antonio, TX. Koch, M., Schnitzler, H.U., 1997. The acoustic startle response in signalling in the neuroendocrine system. J. Neuroendocrinol. rats-circuits mediating evocation, inhibition and potentiation. 20, 743–751. Behav. Brain Res. 89, 35–49. Hurlemann, R., Patin, A., Onur, O.A., Cohen, M.X., Baumgartner, Koenig, J.I., Elmer, G.I., Shepard, P.D., Lee, P.R., Mayo, C., Joy, B., T., Metzler, S., Dziobek, I., Gallinat, J., Wagner, M., Maier, W., Hercher, E., Brady, D.L., 2005. Prenatal exposure to a repeated Kendrick, K.M., 2010. Oxytocin enhances amygdala- variable stress paradigm elicits behavioral and dependent, socially reinforced learning and emotional neuroendocrinological changes in the adult offspring: empathy in humans. J. Neurosci. 30, 4999–5007. potential relevance to schizophrenia. Behav. Brain Res. 156, Jacob, S., u¨ , C.W., Carter, C.S., Leventhal, B.L., Lord, C., Cook Jr., E. 251–261. H., 2007. Association of the oxytocin receptor gene (OXTR) in Kohler, C.G., Walker, J.B., Martin, E.A., Healey, K.M., Moberg, P.J., Caucasian children and adolescents with autism. Neurosci. 2010. Facial emotion perception in schizophrenia: a meta- Lett. 417, 6–9. analytic review. Schizophr. Bull. 36, 1009–1019. Jankowski, M., Bissonauth, V., Gao, L., Gangal, M., Wang, D., Kosfeld, M., Heinrichs, M., Zak, P.J., Fischbacher, U., Fehr, E., 2005. Danalache, B., Wang, Y., Stoyanova, E., Cloutier, G., Blaise, G., Oxytocin increases trust in humans. Nature 435, 673–676. Gutkowska, J., 2010. Anti-inflammatory effect of oxytocin in Kova´cs, G.L., Sarnyai, Z., Szabo, G., 1998. Oxytocin and addiction: rat myocardial infarction. Basic Res. Cardiol. 105, 205–218. a review. Psychoneuroendocrinology 23, 945–962. Jezierski, G., Zehle, S., Bock, J., Braun, K., Gruss, M., 2007. Early Kranzler, H.R., Van Kirk, J., 2001. Efficacy of naltrexone and stress and chronic methylphenidate cross-sensitize acamprosate for alcoholism treatment: a meta-analysis. dopaminergic responses in the adolescent medial prefrontal Alcohol. Clin. Exp. Res. 25, 1335–1341. cortex and nucleus accumbens. J. Neurochem. 103, 2234–2244. Kushner, M.G., Abrams, K., Borchardt, C., 2000. The relationship Jia, R., Tai, F., An, S., Broders, H., Sun, R., 2008a. Neonatal between anxiety disorders and alcohol use disorders: a review manipulation of oxytocin influences the partner preference in of major perspectives and findings. Clin. Psychol. Rev. 20, mandarin voles (Microtus mandarinus). Neuropeptides 42, 149–171. 525–533. Labuschagne, I., Phan, K.L., Wood, A., Angstadt, M., Chua, P., Jia, R., Tai, F.D., An, S.C., Broders, H., Ding, X.L., Kong, Q., Zhao, L., Heinrichs, M., Stout, J.C., Nathan, P.J., 2010. Oxytocin Zhang, H., 2008b. Effects of neonatal oxytocin treatment on attenuates amygdala reactivity to fear in generalized social aggression and neural activities in mandarin voles. Physiol. anxiety disorder. Neuropsychopharmacology 35, 2403–2413. Behav. 95, 56–62. Lee, P.R., Brady, D.L., Shapiro, R.A., Dorsa, D.M., Koenig, J.I., 2005. Jirikowski, G.F., Caldwell, J.D., Stumpf, W.E., Pedersen, C.A., 1990. Social interaction deficits caused by chronic phencyclidine Topography of oxytocinergic estradiol target neurons in the administration are reversed by oxytocin. mouse hypothalamus. Folia Histochem. Cytobiol. 28, 3–9. Neuropsychopharmacology 30, 1883–1894. Jodogne, C., Tirelli, E., Klingbiel, P., Legros, J.-J., 1991. Oxytocin Lee, P.R., Brady, D.L., Shapiro, R.A., Dorsa, D.M., Koenig, J.I., 2007. attenuates tolerance not only to the hypothermic but also to Prenatal stress generates deficits in rat social behavior: the myorelaxant and akinesic effects of ethanol in mice. reversal by oxytocin. Brain Res. 1156, 152–167. Pharmacol. Biochem. Behav. 40, 261–265. Lee, H.J., Macbeth, A.H., Pagani, J.H., Young 3rd, W.S., 2009. Kay, S.R., Fiszbein, A., Opfer, L.A., 1987. The positive and negative Oxytocin: the great facilitator of life. Prog. Neurobiol. 88, syndrome scale (PANSS) for schizophrenia. Schizophr. Bull. 13, 127–151. 261–276. Lee, M.R., Wehring, H.J., McMahon, R.P., Linthicum, J., Cascella, N., Keefe, R.S., Harvey, P.D., 2012. Cognitive impairment in Liu, F., Bellack, A., Buchanan, R.W., Strauss, G.P., Contoreggi, schizophrenia. Handbook of Experimental Pharmacology, vol. C., Kelly, D.L., 2013. Effects of adjunctive intranasal oxytocin 213; 11–37. on olfactory identification and clinical symptoms in 120 brain research 1580 (2014) 102–123

schizophrenia: results from a randomized double blind medications: prescription trends, overall and by physician placebo controlled pilot study. Schizophr. Res. 145, 110–115. specialty. Drug Alcohol Depend. 99, 345–349. Legros, J.J., Gazzotti, C., Carvelli, T., Franchimont, P., Timsit- Mayo-Smith, M.F., 1997. Pharmacological management of alcohol Berthier, M., von Frenckell, R., Ansseau, M., 1992. withdrawal. A meta-analysis and evidence-based practice Apomorphine stimulation of vasopressin- and oxytocin- guideline. American society of addiction medicine working neurophysins. Evidence for increased oxytocinergic and group on pharmacological management of alcohol decreased vasopressinergic function in schizophrenics. withdrawal. J. Am. Med. Assoc. 278, 144–151. Psychoneuroendocrinology 17, 611–617. McCarthy, M.M., McDonald, C.H., Brooks, P.J., Goldman, D., 1996. Linkowski, P., Geenen, V., Kerkhofs, M., Mendlewicz, J., Legros, J.J., An anxiolytic action of oxytocin is enhanced by estrogen in 1984. Cerebrospinal fluid neurophysins in affective illness and the mouse. Physiol. Behav. 60, 1209–1215. in schizophrenia. Eur. Arch. Psychiatry. Neurol. Sci. 234, McCormick, C.M., Mathews, I.Z., 2010. Adolescent development, 162–165. hypothalamic-pituitary-adrenal function, and programming Liu, Y., Wang, Z.X., 2003. Nucleus accumbens oxytocin and of adult learning and memory. Prog. Neuropsychopharmacol. dopamine interact to regulate pair bond formation in female Biol. Psychiatry 34, 756–765. prairie voles. Neuroscience 121, 537–544. McEwen, B.S., 2000. The neurobiology of stress: from serendipity Liu, X., Kawamura, Y., Shimada, T., Otowa, T., Koishi, S., to clinical relevance. Brain Res. 886, 172–189. Sugiyama, T., Nishida, H., Hashimoto, O., Nakagami, R., McEwen, B.S., 2002. Sex, stress and the hippocampus: allostasis, Tochigi, M., Umekage, T., Kano, Y., Miyagawa, T., Kato, N., allostatic load and the aging process. Neurobiol. Aging 23, Tokunaga, K., Sasaki, T., 2010. Association of the oxytocin 921–939. receptor (OXTR) gene polymorphisms with autism spectrum McGregor, I.S., Bowen, M.T., 2012. Breaking the loop: oxytocin as a disorder (ASD) in the Japanese population. J. Hum. Genet. 55, potential treatment for drug addiction. Horm. Behav. 61, 137–141. 331–339. Loup, F., Tribollet, E., Dubois-Dauphin, M., Dreifuss, J.J., 1991. McNair, D., Lorr, M., Droppleman, L., 1971. In: BITS manual for the Localization of high-affinity binding sites for oxytocin and Profile of Mood States. Educational and Industrial Testing vasopressin in the human brain. An autoradiographic study. Service, San Diego, CA. Brain Res. 555, 220–232. Melis, M.R., Melis, T., Cocco, C., Succu, S., Sanna, F., Pillolla, G., Loup, F., Tribollet, E., Dubois-Dauphin, M., Pizzolato, G., Dreifuss, Boi, A., Ferri, G.L., Argiolas, A., 2007. Oxytocin injected into the J.J., 1989. Localization of oxytocin binding sites in the human ventral tegmental area induces penile erection and increases brainstem and upper spinal cord: an autoradiographic study. extracellular dopamine in the nucleus accumbens and Brain Res. 500, 223–230. paraventricular nucleus of the hypothalamus of male rats. Lovinger, D.M., Roberto, M., 2013. Synaptic effects induced by Eur. J. Neurosci. 26, 1026–1035. alcohol. Curr. Top. Behav. Neurosci. 13, 31–86. Meyer-Lindenberg, A., 2010. From maps to mechanisms through MacDonald, K., MacDonald, T.M., 2010. The peptide that binds: a neuroimaging of schizophrenia. Nature 468, 194–202. systematic review of oxytocin and its prosocial effects in Meyer-Lindenberg, A., Domes, G., Kirsch, P., Heinrichs, M., 2011. humans. Harv. Rev. Psychiatry 18, 1–21. Oxytocin and vasopressin in the human brain: social MacDonald, E., Dadds, M.R., Brennan, J.L., Williams, K., Levy, F., neuropeptides for translational medicine. Nat. Rev. Neurosci. Cauchi, A.J., 2011. A review of safety, side-effects and 12, 524–538. subjective reactions to intranasal oxytocin in human Michelini, L.C., 2007. Differential effects of vasopressinergic and research. Psychoneuroendocrinology 36, 1114–1126. oxytocinergic pre-autonomic neurons on circulatory control: MacDonald, K., Feifel, D., 2012. Oxytocin in schizophrenia: a reflex mechanisms and changes during exercise. Clin. Exp. review of evidence for its therapeutic effects. Acta Pharmacol. Physiol. 34, 369–376. Neuropsychiatr. 24, 130–146. Mı¨kolajczak, M., Gross, J.J., Lane, A., Corneille, O., de Timary, P., Macdonald, K., Feifel, D., 2013. Helping oxytocin deliver: Luminet, O., 2010a. Oxytocin makes people trusting, not considerations in the development of oxytocin-based gullible. Psychol. Sci. 21, 1072–1074. therapeutics for brain disorders. Front. Neurosci. 7, 35. Mıkolajczak, M., Pinon, N., Lane, A., de Timary, P., Luminet, O., Mai, J.K., Berger, K., Sofroniew, M.V., 1993. Morphometric ¨ evaluation of neurophysin-immunoreactivity in the human 2010b. Oxytocin not only increases trust when money is at brain: pronounced inter-individual variability and evidence stake, but also when confidential information is in the for altered staining patterns in schizophrenia. J. Hirnforsch. balance. Biol. Psychol. 85, 182–184. 34, 133–154. Moberg, P.J., Turetsky, B.I., 2003. Scent of a disorder: olfactory Malaspina, D., Coleman, E., 2003. Olfaction and social drive in functioning in schizophrenia. Curr. Psychiatry Rep. 5, 311–319. schizophrenia. Arch. Gen Psychiatry 60, 578–584. Moberg, P.J., Arnold, S.E., Doty, R.L., Gur, R.E., Balderston, C.C., Malik, A.I., Zai, C.C., Abu, Z., Nowrouzi, B., Beitchman, J.H., 2012. Roalf, D.R., Gur, R.C., Kohler, C.G., Kanes, S.J., Siegel, S.J., The role of oxytocin and oxytocin receptor gene variants in Turetsky, B.I., 2006. Olfactory functioning in schizophrenia: childhood-onset aggression. Genes Brain Behav. 11, 545–551. relationship to clinical, neuropsychological, and volumetric Mantella, R.C., Vollmer, R.R., Li, X., Amico, J.A., 2003. Female MRI measures. J. Clin. Exp. Neuropsychol. 28, 1444–1461. oxytocin-deficient mice display enhanced anxiety-related Modabbernia, A., Rezaei, F., Salehi, B., Jafarinia, M., Ashrafi, M., behavior. Endocrinology 144, 2291–2296. Tabrizi, M., Hosseini, S.M., Tajdini, M., Ghaleiha, A., Mantella, R.C., Vollmer, R.R., Rinaman, L., Li, X., Amico, J.A., 2004. Akhondzadeh, S., 2013. Intranasal oxytocin as an adjunct to Enhanced corticosterone concentrations and attenuated Fos risperidone in patients with schizophrenia: an 8-week, expression in the medial amygdala of female oxytocin randomized, double-blind, placebo-controlled study. CNS knockout mice exposed to psychogenic stress. Am J. Physiol. Drugs 27, 57–65. Regul. Integr. Comp. Physiol. 287, R1494–R1504. Moghaddam, B., 2002. Stress activation of glutamate Marinelli, M., Rudick, C.N., Hu, X.T., White, F.J., 2006. Excitability neurotransmission in the prefrontal cortex: implications for of dopamine neurons: modulation and physiological dopamine-associated psychiatric disorders. Biol. Psychiatry consequences. CNS Neurol. Disord. Drug Targets 5, 79–97. 51, 775–787. Mark, T.L., Kassed, C.A., Vandivort-Warren, R., Levit, K.R., Montag, C., Brockmann, E.-M., Lehmann, A., Mu¨ ller, D.J., Rujescu, Kranzler, H.R., 2009. Alcohol and opioid dependence D., Gallinat, J., 2012. Association between oxytocin receptor brain research 1580 (2014) 102–123 121

gene polymorphisms and self-rated ‘empathic concern’ in Petrovic, P., Kalisch, R., Singer, T., Dolan, R.J., 2008. Oxytocin schizophrenia. PloS One 7, e51882. attenuates affective evaluations of conditioned faces and Montag, C., Brockmann, E.M., Bayerl, M., Rujescu, D., Muller, D.J., amygdala activity. J. Neurosci. 28, 6607–6615. Gallinat, J., 2013. Oxytocin and oxytocin receptor gene Phillips, M.L., Drevets, W.C., Rauch, S.L., Lane, R., 2003. polymorphisms and risk for schizophrenia: a case-control Neurobiology of emotion perception II: implications for major study. World J. Biol. Psychiatry 14, 500–508. psychiatric disorders. Biol. Psychiatry 54, 515–528. Myrick, H., Malcolm, R., Randall, P.K., Boyle, E., Anton, R.F., Becker, Pinkham, A.E., Penn, D.L., Perkins, D.O., Lieberman, J., 2003. H.C., Randall, C.L., 2009. A double-blind trial of gabapentin Implications for the neural basis of social cognition for the versus lorazepam in the treatment of alcohol withdrawal. study of schizophrenia. Am. J. Psychiatry 160, 815–824. Alcohol. Clin. Exp. Res. 33, 1582–1588. Pinkham, A.E., Penn, D.L., 2006. Neurocognitive and social Neumann, I.D., Wigger, A., Torner, L., Holsboer, F., Landgraf, R., cognitive predictors of interpersonal skill in schizophrenia. 2000. Brain oxytocin inhibits basal and stress-induced activity Psychiatry Res. 143, 167–178. of the hypothalamo-pituitary-adrenal axis in male and female Pobbe, R.L., Pearson, B.L., Blanchard, D.C., Blanchard, R.J., 2012a. rats: partial action within the paraventricular nucleus. J. Oxytocin receptor and Mecp2 308/Y knockout mice exhibit Neuroendocrinol. 12, 235–243. altered expression of autism-related social behaviors. Physiol. Ninan, I., 2011. Oxytocin suppresses basal glutamatergic Behav. 107, 641–648. transmission but facilitates activity-dependent synaptic Pobbe, R.L., Pearson, B.L., Defensor, E.B., Bolivar, V.J., Young 3rd, potentiation in the medial prefrontal cortex. J. Neurochem. W.S., Lee, H.J., Blanchard, D.C., Blanchard, R.J., 2012b. Oxytocin 119, 324–331. receptor knockout mice display deficits in the expression of Norman, G.J., Cacioppo, J.T., Morris, J.S., Malarkey, W.B., Berntson, autism-related behaviors. Horm. Behav. 61, 436–444. G.G., Devries, A.C., 2011. Oxytocin increases autonomic Pucilowski, O., Kostowski, W., Trzaskowska, E., 1985. The effect of cardiac control: moderation by loneliness. Biol. Psychol. 86, oxytocin and fragment (MIF-I) on the development of 174–180. tolerance to hypothermic and hypnotic action of ethanol in Numan, M., Insel, T.R., 2003. In: The Neurobiology of Parental the rat. Peptides 6, 7–10. Behavior. Springer Verlag, New York. Qi, J., Yang, J.Y., Song, M., Li, Y., Wang, F., Wu, C.F., 2008. Inhibition Overall, J.E., Gorham, D.R., 1962. The Brief Psychiatric Rating by oxytocin of methamphetamine-induced hyperactivity Scale. Psychol. Rep. 10, 799–812. related to dopamine turnover in the mesolimbic region in Overstreet, D.H., Knapp, D.J., Breese, G.R., 2007. Drug challenges mice. Naunyn-Schmiedeberg’s Arch. Pharmacol. 376, 441–448. reveal differences in mediation of stress facilitation of Qiao, H., Noda, Y., Kamei, H., Nagai, T., Furukawa, H., Miura, H., voluntary alcohol drinking and withdrawal-induced anxiety in Kayukawa, Y., Ohta, T., Nabeshima, T., 2001. Clozapine, but alcohol-preferring P rats. Alcohol. Clin. Exp. Res. 31, not haloperidol, reverses social behavior deficit in mice during 1473–1481. withdrawal from chronic phencyclidine treatment. Patisaul, H.B., Scordalakes, E., Young, L.J., Rissman, E.F., 2003. Neuroreport 12, 11–15. Oxytocin, but not oxytocin receptor, is regulated by oestrogen Riem, M.M., van, I.M.H., Tops, M., Boksem, M.A., Rombouts, S.A., receptor β in the female mouse hypothalamus. J. Bakermans-Kranenburg, M.J., 2012. No laughing matter: Neuroendocrinol. 15, 787–793. intranasal oxytocin administration changes functional brain Pava, M.J., Woodward, J.J., 2012. A review of the interactions connectivity during exposure to infant laughter. between alcohol and the endocannabinoid system: Neuropsychopharmacology 37, 1257–1266. implications for alcohol dependence and future directions for Rigter, H., Dortmans, C., Crabbe Jr., J.C., 1980. Effects of peptides research. Alcohol 46, 185–204. related to neurohypophyseal hormones on ethanol tolerance. Pedersen, C.A., Caldwell, J.D., Jirikowski, G.F., Insel, T.R., 1992. Pharmacol. Biochem. Behav. 13 (Suppl. 1), S285–S290. Oxytocin in maternal, sexual, and social behaviors. Ann. N. Y. Rilling, J.K., DeMarco, A.C., Hackett, P.D., Thompson, R., Ditzen, B., Acad. Sci., 652. Patel, R., Pagnoni, G., 2012. Effects of intranasal oxytocin and Pedersen, C.A., Gibson, C.M., Rau, S.W., Salimi, K., Smedley, K.L., vasopressin on cooperative behavior and associated brain Casey, R.L., Leserman, J., Jarskog, L.F., Penn, D.L., 2011. activity in men. Psychoneuroendocrinology 37, 447–461. Intranasal oxytocin reduces psychotic symptoms and Ring, R.H., Malberg, J.E., Potestio, L., Ping, J., Boikess, S., Luo, B., improves theory of mind and social perception in Schechter, L.E., Rizzo, S., Rahman, Z., Rosenzweig-Lipson, S., schizophrenia. Schizophr. Res. 132, 50–53. 2006. Anxiolytic-like activity of oxytocin in male mice: Pedersen, C.A., 2013. Oxytocin regulation of maternal behavior. behavioral and autonomic evidence, therapeutic implications. From rodent to humans. In: Choleris, E., Pfaff, D.W., Kavaliers., Psychopharmacology (Berl.) 185, 218–225. M. (Eds.), Oxytocin, Vasopressin and Related Peptides in the Roberto, M., Gilpin, N.W., Siggins, G.R., 2012. The central Regulation of Behavior. Cambridge University Press, amygdala and alcohol: role of gamma-aminobutyric acid, Cambridge, UK, pp. 148–182. glutamate, and neuropeptides. Cold Spring Harb. Perspect. Pedersen, C.A., Smedley, K.L., Leserman, J., Jarskog, L.F., Rau, S.W., Med. 2, a012195. Kampov-Polevoi, A., Casey, R.L., Fender, T., Garbutt, J.C., 2013. Rubin, L.H., Carter, C.S., Drogos, L., Jamadar, R., Pournajafi- Intranasal oxytocin blocks alcohol withdrawal in human Nazarloo, H., Sweeney, J.A., Maki, P.M., 2011. Sex-specific subjects. Alcohol. Clin. Exp. Res. 37, 484–489. associations between peripheral oxytocin and emotion Penn, D.L., Addington, J., Pinkham, A., 2006. Social cognitive perception in schizophrenia. Schizophr. Res. 130, 266–270. impairments. In: Lieberman, J.A., Stroup, T.S., Perkins, D.O. Rubin, L.H., Carter, C.S., Drogos, L., Pournajafi-Nazarloo, H., (Eds.), The American Psychiatric Publishing Textbook of Sweeney, J.A., Maki, P.M., 2010. Peripheral oxytocin is Schizophrenia. American Psychiatric Publishing, Inc., associated with reduced symptom severity in schizophrenia. Arlington, VA, USA, pp. 261–274. Schizophr. Res. 124, 13–21. Penn, D.L., Keefe, R.S., Davis, S.M., Meyer, P.S., Perkins, D.O., Sams-Dodd, F., 1999. Phencyclidine in the social interaction test: Losardo, D., Lieberman, J.A., 2009. The effects of antipsychotic an animal model of schizophrenia with face and predictive medications on emotion perception in patients with chronic validity. Rev. Neurosci. 10, 59–90. schizophrenia in the CATIE trial. Schizophr. Res. 115, 17–23. Sarnyai, Z., Szabo, G., Kova´cs, G.L., Telegdy, G., 1990. Oxytocin Penn, D.L., Sanna, L.J., Roberts, D.L., 2008. Social cognition in attenuates the cocaine-induced exploratory hyperactivity in schizophrenia: an overview. Schizophr. Bull. 34, 408–411. mice. Neuroreport 1, 200–202. 122 brain research 1580 (2014) 102–123

Sarnyai, Z., Kova´cs, G.L., 1994. Role of oxytocin in the Szabo´, G., Kova´cs, G.L., Telegdy, G., 1987. Effects of neuroadaptation to drugs of abuse. neurohypophyseal peptide hormones on alcohol dependence Psychoneuroendocrinology 19, 85–117. and withdrawal. Alcohol 22, 71–74. Sasayama, D., Hattori, K., Teraishi, T., Hori, H., Ota, M., Yoshida, Szabo´, G., Kova´cs, G.L., Telegdy, G., 1989. Intraventricular S., Arima, K., Higuchi, T., Amano, N., Kunugi, H., 2012. administration of neurohypophyseal hormones interferes Negative correlation between cerebrospinal fluid oxytocin with the development of tolerance to ethanol. Acta Physiol. levels and negative symptoms of male patients with Hung. 73, 97–103. schizophrenia. Schizophr. Res. 139, 201–206. Szczepanska-Sadowska, E., 2008. Role of neuropeptides in central Schmitt, A., Hasan, A., Gruber, O., Falkai, P., 2011. Schizophrenia control of cardiovascular responses to stress. J. Physiol. as a disorder of disconnectivity. Eur. Arch. Psychiatry. Clin. Pharmacol. 59 (Suppl. 8), S61–S89. Neurosci. 261 (Suppl. 2), S150–S154. Szeto, A., Nation, D.A., Mendez, A.J., Dominguez-Bendala, J., Sergi, M.J., Green, M.F., Widmark, C., Reist, C., Erhart, S., Braff, D. Brooks, L.G., Schneiderman, N., McCabe, P.M., 2008. Oxytocin L., Kee, K.S., Marder, S.R., Mintz, J., 2007. Social cognition and attenuates NADPH-dependent superoxide activity and IL-6 neurocognition: effects of risperidone, olanzapine, and secretion in macrophages and vascular cells. Am. J. Physiol. haloperidol. Am. J. Psychiatry 164, 1585–1592. Endocrinol. Metab. 295, E1495–E1501. Shahrokh, D.K., Zhang, T.Y., Diorio, J., Gratton, A., Meaney, M.J., Takayanagi, Y., Yoshida, M., Bielsky, I.F., Ross, H.E., Kawamata, M., 2010. Oxytocin-dopamine interactions mediate variations in Onaka, T., Yanagisawa, T., Kimura, T., Matzuk, M.M., Young, L.J., maternal behavior in the rat. Endocrinology 151, 2276–2286. Nishimori, K., 2005. Pervasive social deficits, but normal Shamay-Tsoory, S.G., Fischer, M., Dvash, J., Harari, H., Perach- parturition, in oxytocin receptor-deficient mice. Proc. Natl. Acad. Bloom, N., Levkovitz, Y., 2009. Intranasal administration of Sci. USA 102, 16096–16101. oxytocin increases envy and schadenfreude (gloating). Biol. Takemura, M., Kimura, T., Nomura, S., Makino, Y., Inoue, T., Psychiatry 66, 864–870. Kikuchi, T., Kubota, Y., Tokugawa, Y., Nobunaga, T., Kamiura, S., Shapiro, L.E., Insel, T.R., 1989. Ontogeny of oxytocin receptors in Onoue, H., Azuma, C., Saji, F., Kitamura, Y., Tanizawa, O., 1994. rat forebrain: a quantitative study. Synapse 4, 259–266. Expression and localization of human oxytocin receptor mRNA Shughrue, P.J., Dellovade, T.L., Merchenthaler, I., 2002. Estrogen and its protein in chorion and decidua during parturition. J. modulates oxytocin gene expression in regions of the rat Clin. Investig. 93, 2319–2323. supraoptic and paraventricular nuclei that contain estrogen Theodoridou, A., Rowe, A.C., Penton-Voak, I.S., Rogers, P.J., 2009. receptor-beta. Prog. Brain Res. 139, 15–29. Oxytocin and social perception: oxytocin increases perceived Smith, A.L., Freeman, S.M., Stehouwer, J.S., Inoue, K., Voll, R.J., facial trustworthiness and attractiveness. Horm. Behav. 56, Young, L.J., Goodman, M.M., 2012. Synthesis and evaluation of 128–132. C-11, F-18 and I-125 small molecule radioligands for detecting Thompson, K.N., Berger, G., Phillips, L.J., Komesaroff, P., Purcell, oxytocin receptors. Bioorg. Med. Chem. 20, 2721–2738. R., McGorry, P.D., 2007a. HPA axis functioning associated with Souza, R.P., de Luca, V., Meltzer, H.Y., Lieberman, J.A., Kennedy, J. transition to psychosis: combined DEX/CRH test. J. Psychiatr. L., 2010. Schizophrenia severity and clozapine treatment Res. 41, 446–450. outcome association with oxytocinergic genes. Int. J. Thompson, K.N., Phillips, L.J., Komesaroff, P., Yuen, H.P., Wood, S.J., Neuropsychopharmacol. 13, 793–798. Pantelis, C., Velakoulis, D., Yung, A.R., McGorry, P.D., 2007b. Sripada, C.S., Phan, K.L., Labuschagne, I., Welsh, R., Nathan, P.J., Stress and HPA-axis functioning in young people at ultra high Wood, A.G., 2013. Oxytocin enhances resting-state risk for psychosis. J. Psychiatr. Res. 41, 561–569. connectivity between amygdala and medial frontal cortex. Int. Toloczko, D.M., Young, L., Insel, T.R., 1997. Are there oxytocin J. Neuropsychopharmacol. 16, 255–260. receptors in the primate brain?. Ann. N. Y. Acad. Sci. 807, Stephan, K.E., Baldeweg, T., Friston, K.J., 2006. Synaptic plasticity 506–509. and dysconnection in schizophrenia. Biol. Psychiatry 59, Tomizawa, K., Iga, N., Lu, Y.F., Moriwaki, A., Matsushita, M., Li, S. 929–939. T., Miyamoto, O., Itano, T., Matsui, H., 2003. Oxytocin improves Stephan, K.E., Friston, K.J., Frith, C.D., 2009. Dysconnection in long-lasting spatial memory during motherhood through MAP schizophrenia: from abnormal synaptic plasticity to failures kinase cascade. Nat. Neurosci. 6, 384–390. Tribollet, E., Charpak, S., Schmidt, A., Dubois-Dauphin, M., of self-monitoring. Schizophr. Bull. 35, 509–527. Strauss, G.P., Allen, D.N., Ross, S.A., Duke, L.A., Schwartz, J., 2010. Dreifuss, J.J., 1989. Appearance and transient expression of Olfactory hedonic judgment in patients with deficit syndrome oxytocin receptors in fetal, infant, and peripubertal rat brain studied by autoradiography and electrophysiology. J. Neurosci. schizophrenia. Schizophr. Bull. 36, 860–868. Succu, S., Sanna, F., Melis, T., Boi, A., Argiolas, A., Melis, M.R., 9, 1764–1773. Tribollet, E., Dubois-Dauphin, M., Dreifuss, J.J., Barberis, C., Jard, 2007. Stimulation of dopamine receptors in the S., 1992. Oxytocin receptors in the central nervous system. paraventricular nucleus of the hypothalamus of male rats Distribution, development, and species differences. Ann. N. Y. induces penile erection and increases extra-cellular dopamine Acad. Sci. 652, 29–38. in the nucleus accumbens: involvement of central oxytocin. Uvna¨s-Moberg, K., Alster, P., Svensson, T.H., 1992. Amperozide Neuropharmacology 52, 1034–1043. and clozapine but not haloperidol or raclopride increase the Sullivan, J.T., Sykora, K., Schneiderman, J., Naranjo, C.A., Sellers, secretion of oxytocin in rats. Psychopharmacology (Berl.) 109, E.M., 1989. Assessment of alcohol withdrawal: the revised 473–476. Clinical Institute Withdrawal Assessment for Alcohol Scale Vadlamudi, S., Pedersen, C.A., Amico, J.A., Breese, G.R., Knapp, D.J., (CIWA-Ar). Br. J. Addict. 84, 1353–1357. 2004. Neuroadaptation to alcohol and morphine is altered in Swerdlow, N.R., Geyer, M.A., Braff, D.L., 2001. Neural circuit oxytocin knockout mice, Society for Neuroscience Abstract regulation of prepulse inhibition of startle in the rat: current Viewer/Itinerary Planner. knowledge and future challenges. Psychopharmacology (Berl.) van Os, J., Krabbendam, L., Myin-Germeys, I., Delespaul, P., 2005. 156, 194–215. The schizophrenia enviroment. Curr. Opin. Psychiatry 18, Szabo´, G., Kova´cs, G.L., Szekeli, S., Telegdy, G., 1985. The effects of 141–145. neurohypophyseal hormones on tolerance to the hypothermic van Os, J., Kenis, G., Rutten, B.P., 2010. The environment and effect of ethanol. Alcohol 2, 567–574. schizophrenia. Nature 468, 203–212. brain research 1580 (2014) 102–123 123

Vazquez, D.M., Akil, H., 1993. Pituitary-adrenal response to ether Wsol, A., Cudnoch-Jedrzejewska, A., Szczepanska-Sadowska, E., vapor in the weanling animal: characterization of the Kowalewski, S., Puchalska, L., 2008. Oxytocin in the inhibitory effect of glucocorticoids on adrenocorticotropin cardiovascular responses to stress. J. Physiol. Pharmacol. 59 secretion. Pediatr. Res. 34, 646–653. (Suppl. 8), S123–S127. Walker, E.F., Diforio, D., 1997. Schizophrenia: a neural diathesis- Wu, S., Jia, M., Ruan, Y., Liu, J., Guo, Y., Shuang, M., Gong, X., stress model. Psychol. Rev. 104, 667–685. Zhang, Y., Yang, X., Zhang, D., 2005. Positive association of the Walker, E., Mittal, V., Tessner, K., 2008. Stress and the oxytocin receptor gene (OXTR) with autism in the Chinese hypothalamic pituitary adrenal axis in the developmental Han population. Biol. Psychiatry 58, 74–77. course of schizophrenia. Ann. Rev. Clin. Psychol. 4, 189–216. Yager, J.A., Ehmann, T.S., 2006. Untangling social function and Walker, E.F., Sabuwalla, Z., Huot, R., 2004. Pubertal social cognition: a review of concepts and measurement. neuromaturation, stress sensitivity, and psychopathology. Psychiatry 69, 47–68. Dev. Psychopathol. 16, 807–824. Yang, J.Y., Qi, J., Han, W.Y., Wang, F., Wu, C.F., 2010. Inhibitory role Wechsler, D., 1997. Wechsler Adult Intelligence Scale Third of oxytocin in psychostimulant-induced psychological Edition (WAIS-III). Psychological Corporation, San Antonio. dependence and its effects on dopaminergic and Wee, S., Koob, G.F., 2010. The role of the dynorphin-kappa opioid glutaminergic transmission. Acta Pharmacol. Sin. 31, system in the reinforcing effects of drugs of abuse. 1071–1074. Psychopharmacology (Berl.) 210, 121–135. Yoshida, M., Takayanagi, Y., Inoue, K., Kimura, T., Young, L.J., Windle, R.J., Shanks, N., Lightman, S.L., Ingram, C.D., 1997. Central oxytocin administration reduces stress-induced Onaka, T., Nishimori, K., 2009. Evidence that oxytocin exerts corticosterone release and anxiety behavior in rats. anxiolytic effects via oxytocin receptor expressed in Endocrinology 138, 2829–2834. serotonergic neurons in mice. J. Neurosci. 29, 2259–2271. Wittfoth-Schardt, D., Grunding, J., Wittfoth, M., Lanfermann, H., Zak, P.J., Kurzban, R., Matzner, W.T., 2005. Oxytocin is associated Heinrichs, M., Domes, G., Buchheim, A., Gundel, H., Waller, C., with human trustworthiness. Horm. Behav. 48, 522–527. 2012. Oxytocin modulates neural reactivity to children’s faces Zink, C.F., Meyer-Lindenberg, A., 2012. Human neuroimaging of as a function of social salience. Neuropsychopharmacology oxytocin and vasopressin in social cognition. Horm. Behav. 61, 37, 1799–1807. 400–409.